Steinberg MR816 vs Apogee Ensamble - The Definative Answer please! (No specualtion)

[smoovemode]

I used to be a windows dude and I'm still a windows technician. I switched to Mac because of the driver issues I was having with various pieces of musical equipment. If it is on windows, I would not blame the device. I found many windows boxes were simply not made for musical equipment in mind. More so the laptops! All I know is when I plug my gear into a Mac...it works!

[adrianww]

Originally Posted by kaiO

Thank you for your reply. On Windows or Mac? What kind of the driver issue? I can't find any other post regarding the instability of Steinberg driver, except the issue when chaining 2 or more MR816s at the same time.

This was on an old Windows XP (SP3, 32-bit) machine. Don't get me wrong, I did get the n12 working and used it for a few years. Loved the hardware, but I couldn't run it at anything lower than 256 or 512 samples latency (15-20ms round trip at 44.1kHz 16-bit) without risking clicks, pops and dropouts. On the same machine, a crappy old Alesis iO/14 would happily run down to 128 samples (or even sometimes 64 for simple projects). The ASIO meter in Cubase also ran higher with the n12 than the iO/14. I guess it's possible that the Alesis just had a much better FireWire chipset in it than the n12, but I wouldn't necessarily bet on that. Also, switching to ASIO4ALL made the iO/14 slightly worse, while it made the n12 slightly better (although you lost use of the Cubase remote functions if you didn't use the Yamaha/Steinberg FW driver). Under the circumstances, I suspect it was the Yamaha/Steinberg driver that has to shoulder a large part of the blame there. It's certainly made me wary of choosing Yamaha/Steinberg hardware in the future. (Well, except for Yamaha instruments, but you know what I mean.)

My newer machine is a 64-bit Win7 box and, so far, it runs my LS56 beautifully, all the way down to the lowest latency that the interface supports (64 samples, around 3ms round trip at 44.1kHz 24-bit). I haven't got around to downloading the Yamaha/Steinberg Win7 driver and trying the n12 on the new machine yet, but I'll probably do that at some point and see how it matches up to the Focusrite.

Of course, all of this is just my experience - it might be typical, it might be some bizarre one-off. Like all anecdotes, it's not statistically significant. Also, it doesn't involve the MR range which are USB rather than Firewire, so who knows - maybe their USB drivers are fine and it's only their FireWire stuff that was problematic. I guess, ultimately, only you will be able to judge whether something really meets your needs and whether you like it. It's just a bummer that you can only really make that call once you've used something for a while and lived with it. (Been there, done that, worn out the T-shirt a few times...)

Actually, having said all that, I was probably being a little unfair in my last post when I said that I don't trust Steinberg hardware drivers "at all". It's more accurate to say that I just don't trust them as much as some others.

[ Edit: Just revisiting this old thread 'cos it popped back up again and realised that I said the MR series were USB devices there. Obviously, I was thinking of the UR series when I said that, the MR's are FireWire aren't they! D'oh! ]

[SoSueMe]

Originally Posted by drumdrum

Mic choice and its impact on a mix always trumps converter and preamp choice. Always.

Dont get hung up on converter choice. Your microphone will make or break a sound. Converters won't.

I own a mr816x, and don't let that other thread fool you. They are amazing pieces of kit (as also are rme and apogee stuff) and all 3 are more than capable of producing a professional record. I own as 816x and have used a ff800 and ensemble before. Both are also stellar pieces of kit, though the ensemble is more expensive. Ff800 only gives you 4 preamps also, though you say that that would be ok for you so I would still consider it. Preamp quality is subjective anyways, and tbh, its got maybe at most 5% impact on the sound anyways. (the test being, player, instrument quality, mic choice and mic position)

Remember, records are made by people, not converters, so dont worry too much about these 3 brands / products. Get one of these (i recommend the 816x if purely from price point), get some decent mics, and most importantly, learn how to use them. The biggest asset to any studio is the engineer...not the gear.

Best of luck!

What he said, except the room is just as important as the mic/position, even with cardioids up close depending on the instrument.

[smoovemode]

My Shure ksm44 comes in today. I'm hoping it's a good match with my LS56

[RebornOnline]

Originally Posted by adrianww

Don't know about the preamps, but when it comes to AD/DA conversion, one of the few (possibly the only?) scientific or objective test of different converters here on GS seems to suggest that the MR's converters are actually very poor - particularly on the AD side.

The test is a loopback, so it emphasises any deterioration in AD or DA.

So it might have amazing AD (which is what we care about), but the DA (when multiplied 10^9 (which is alot)) sounds bad* - then you can get these results.

Similarly, it might have rubbish AD and good DA.

My point is that with this test you can't tell, it's just an overall measurement. This unit could be completely fine (unless you regularly bouncedown tracks by going in and out of the converters).

Just my 2p

*or at least - "looks" bad.

[kaiO]

Originally Posted by adrianww

This was on an old Windows XP (SP3, 32-bit) machine. Don't get me wrong, I did get the n12 working and used it for a few years. Loved the hardware, but I couldn't run it at anything lower than 256 or 512 samples latency (15-20ms round trip at 44.1kHz 16-bit) without risking clicks, pops and dropouts. On the same machine, a crappy old Alesis iO/14 would happily run down to 128 samples (or even sometimes 64 for simple projects). The ASIO meter in Cubase also ran higher with the n12 than the iO/14. I guess it's possible that the Alesis just had a much better FireWire chipset in it than the n12, but I wouldn't necessarily bet on that. Also, switching to ASIO4ALL made the iO/14 slightly worse, while it made the n12 slightly better (although you lost use of the Cubase remote functions if you didn't use the Yamaha/Steinberg FW driver). Under the circumstances, I suspect it was the Yamaha/Steinberg driver that has to shoulder a large part of the blame there. It's certainly made me wary of choosing Yamaha/Steinberg hardware in the future. (Well, except for Yamaha instruments, but you know what I mean.)

My newer machine is a 64-bit Win7 box and, so far, it runs my LS56 beautifully, all the way down to the lowest latency that the interface supports (64 samples, around 3ms round trip at 44.1kHz 24-bit). I haven't got around to downloading the Yamaha/Steinberg Win7 driver and trying the n12 on the new machine yet, but I'll probably do that at some point and see how it matches up to the Focusrite.

Of course, all of this is just my experience - it might be typical, it might be some bizarre one-off. Like all anecdotes, it's not statistically significant. Also, it doesn't involve the MR range which are USB rather than Firewire, so who knows - maybe their USB drivers are fine and it's only their FireWire stuff that was problematic. I guess, ultimately, only you will be able to judge whether something really meets your needs and whether you like it. It's just a bummer that you can only really make that call once you've used something for a while and lived with it. (Been there, done that, worn out the T-shirt a few times...)

Actually, having said all that, I was probably being a little unfair in my last post when I said that I don't trust Steinberg hardware drivers "at all". It's more accurate to say that I just don't trust them as much as some others.

I saw a few other posts regarding the incompatible onboard VIA firewire chipset issue (Asus, etc). It seems like spending extra $20 for a firefire PCI card with TI chip is a good idea. In Nuendo forum, someone actually could get 64 samples with 90% Asio load with no dropout. If it's true, it's even better than RME firewire IO, although it's not better than their PCIe interfaces.

[adrianww]

Originally Posted by kaiO

I saw a few other posts regarding the incompatible onboard VIA firewire chipset issue (Asus, etc). It seems like spending extra $20 for a firefire PCI card with TI chip is a good idea. In Nuendo forum, someone actually could get 64 samples with 90% Asio load with no dropout. If it's true, it's even better than RME firewire IO, although it's not better than their PCIe interfaces.

I've been using a PCI FireWire card with a TI chipset throughout - on both the old system and the new one. The other drivers still outperformed the Yamaha/Steinberg one. But again, that's just my experience - it could be different for other folks.

[adrianww]

Originally Posted by RebornOnline

The test is a loopback, so it emphasises any deterioration in AD or DA.

So it might have amazing AD (which is what we care about), but the DA (when multiplied 10^9 (which is alot)) sounds bad* - then you can get these results.

Similarly, it might have rubbish AD and good DA.

My point is that with this test you can't tell, it's just an overall measurement. This unit could be completely fine (unless you regularly bouncedown tracks by going in and out of the converters).

Just my 2p

*or at least - "looks" bad.

I think one of the reasons the test was set up was precisely to consider the situation you describe - bouncing out and back in to use external hardware FX or OTB summing or whatever. (I apologise in advance and I'm sure that someone will come along and correct me in that if I'm misreading or misremembering things.)

Also, I think you're right about the AD vs. DA performance issue. I think the test has also been tried going out of the box via some other route and back in through the MR. In that case, the MR's performance seemed to improve. There's also a thread on GS somewhere about which converter chips are used in various interfaces - looking on there it seems like the MR series actually use the signal path from the n12 (or something close to it) but with a lower-quality multi-channel DA stage handling the outputs. If that's true, it would also go some way to explaining the results.

[kaiO]

Originally Posted by adrianww

Also, I think you're right about the AD vs. DA performance issue. I think the test has also been tried going out of the box via some other route and back in through the MR. In that case, the MR's performance seemed to improve. There's also a thread on GS somewhere about which converter chips are used in various interfaces - looking on there it seems like the MR series actually use the signal path from the n12 (or something close to it) but with a lower-quality multi-channel DA stage handling the outputs. If that's true, it would also go some way to explaining the results.

I checked various converter chip threads. MR816 AD and DA chips are not multi channel ones. 2 ch chips. They are AKM's second highest grade converter chips. They are probably more expensive ones that used in n12, (but not sure, since there is no info about n12 chips.)

I think you are right about AD of MR816. MR816 line inputs are actually padded mic pres, so it has an obvious disadvantage for loop back test result. I know Yamaha's mic pres are surprisingly good, but it is rather musical than transparent.

Do you really think it's not worth? If so, I would like to hear what is your current recommendation for under 1K multi channel converter.

[adrianww]

The thread about the converter chips is in the Geekslutz forum here. Someone has checked the service manuals for the n12 and the MR816. They both use identical 2-ch AD units (AK5385) but the n12 uses multiple 2-ch DA chips (AK4393) while the MR816 uses an 8-channel codec (AK4358) for DA. The noise/THD and dynamic range performance of the 8-channel DA is not as good as the 2-channel DA's. (It was originally thought that the n12 and MR816 converters were identical, but when someone checked the service manuals, they found the difference in the DA side.)

I don't know for sure whether the n12 and MR816 preamps are exactly the same, but given that their published descriptions and specifications are pretty much identical, even down to the gain range that is available, it's not a bad assumption.

Having said all of that, I never had a problem with my n12 in terms of its sound and I don't think there's any big issue with the MR series either. I know that it doesn't seem to have done well in the famous (infamous?) AD-DA shoot out test thread, but I'm sure it works fine and is a good interface that can be used to create/record great music. I only moved away from the n12 because I wasn't completely happy with the drivers. I've still got it though and I still like it enough that I might bring it back into use just as a mixer or even as an alternate/backup interface.

I honestly think that, for things that are roughly the same price (within, say, a couple of hundred or even a few hundred dollars/pounds/euro of one another), most interfaces are going to be roughly in the same ball park nowadays when it comes to sound quality. The technology is mature enough that, so long as the manufacturer doesn't make a complete screw-up of something, other things (e.g. mic choice, room acoustics, performance on the day) are probably going to have far more impact on your end results than your AD-DA stages. So the choice of interface becomes more about whether the number and type of I/O suits what you want to do, whether there are any particular bells and whistles you want (e.g. DSP effects, bypass-able pres) and whether it will fit in well with the way you work (PC vs Mac, your chosen DAW, etc.)

But I don't claim to be any great expert on this - I'm not a pro engineer, just an ex-computer industry guy and part-time keyboard player who likes to tinker with this stuff at home. It's just my take on things.

[kaiO]

Originally Posted by adrianww

The thread about the converter chips is in the Geekslutz forum here. Someone has checked the service manuals for the n12 and the MR816. They both use identical 2-ch AD units (AK5385) but the n12 uses multiple 2-ch DA chips (AK4393) while the MR816 uses an 8-channel codec (AK4358) for DA. The noise/THD and dynamic range performance of the 8-channel DA is not as good as the 2-channel DA's. (It was originally thought that the n12 and MR816 converters were identical, but when someone checked the service manuals, they found the difference in the DA side.)

I don't know for sure whether the n12 and MR816 preamps are exactly the same, but given that their published descriptions and specifications are pretty much identical, even down to the gain range that is available, it's not a bad assumption.

Having said all of that, I never had a problem with my n12 in terms of its sound and I don't think there's any big issue with the MR series either. I know that it doesn't seem to have done well in the famous (infamous?) AD-DA shoot out test thread, but I'm sure it works fine and is a good interface that can be used to create/record great music. I only moved away from the n12 because I wasn't completely happy with the drivers. I've still got it though and I still like it enough that I might bring it back into use just as a mixer or even as an alternate/backup interface.

I honestly think that, for things that are roughly the same price (within, say, a couple of hundred or even a few hundred dollars/pounds/euro of one another), most interfaces are going to be roughly in the same ball park nowadays when it comes to sound quality. The technology is mature enough that, so long as the manufacturer doesn't make a complete screw-up of something, other things (e.g. mic choice, room acoustics, performance on the day) are probably going to have far more impact on your end results than your AD-DA stages. So the choice of interface becomes more about whether the number and type of I/O suits what you want to do, whether there are any particular bells and whistles you want (e.g. DSP effects, bypass-able pres) and whether it will fit in well with the way you work (PC vs Mac, your chosen DAW, etc.)

But I don't claim to be any great expert on this - I'm not a pro engineer, just an ex-computer industry guy and part-time keyboard player who likes to tinker with this stuff at home. It's just my take on things.

OK, Thanks a lot.

[smoovemode]

I still would like to know how good the preamps on the MR816 and the Ensemble stack up against something like an ISA 428 or other dedicated preamps?

[mitchiemasha]

So, I'm still non the wiser. Needed a new interface since my emu 0404 wont work with my latest pc. That was like 6-8 months ago. Getting sick of reading now.

[adrianww]

Originally Posted by mitchiemasha

So, I'm still non the wiser. Needed a new interface since my emu 0404 wont work with my latest pc. That was like 6-8 months ago. Getting sick of reading now.

OK - quick pop quiz: how many inputs/outputs do you want; are you PC or Mac based (and which OS version); how do you prefer to connect up (USB, FW, etc.) and how much do you want to spend?

[mitchiemasha]

Prefer Firewire or pci express? Not usb, My Virus TI runs on usb and sharing the ports can mess things up. PC, windows 7, 64 bit, running cubase 5. Min 6in 6out, seriously low latecy. All the other trimmings I have no need for. Need the AD conversion to be the best available. Not fussed on amount of Preamps. As for price, I just don't want to be paying more for features I don't need.

[dkelley]

my quick thoughts:

the loopback tests are perfectly valid but don't test clock jitter. according to one user here testing clock jitter is vital, but many of us disagree. the sound quality of hte conversion alone is still tested, with either a/d or d/a degradation affecting the summed outcome. What is invalid about that?

loopback tests are INSANELY simple to operate with the same software and a simple cable. variation in the cable will make NO difference unless it's a very long cable of very poor quality. the person who claimed it makes a difference doesn't understand how length and electrical characteristics of digital cables only alter the decay time of digital 1 to 0 value changes, and how that is very effectively read by any d/a converter unless it's extremely out of spec and by that time you're hearing very clear issues. It also only occurs with very poor quality cables over long lengths.

But mostly, on the poitn of this thread, the mr816 uses an 8 channel d/a converter who's specs are poorer than any other product I've seen specs for in the past several years. However it's a/d conversion is excellent.

Interestingly, the a/d conversion is the same excellent conversion as on my rather old m-audio delta 1010 card which also has very high quality d/a conversion using multiple 2 channel conversion chips.

I find it most amusing that everyone freaks out about conversion when it repeatedly is proven that some of the best products around have always been great, even if they're from companies that people think ought to not be so great (m-audio for just one example), and for some other products that by design MUST have very poor d/a conversion which is tested and proven but also backed up by the specs of the d/a converter chip itself... well products like that (the mr816) get rave reviews from people who claim it sounds so much better (and yet it has a clear frequency balance that is off from flat compared to most other products tested).

So what matters is, as I've always said, driver reliability and features.

If the worst spec'd product sounds great to many pro ears then just enjoy ALL of the amazing converters out there these days since even the worst are clearly plenty good enough for pro work.

Choose something that fits everyone's needs and enjoy it.

And yes, there is the unknown and untested clock jitter of all of these products still which could technically affect the conversion quality. However anyone who actually knows their circuit design will see that every one of these products uses nearly identical clock crystals and nearly every one does the same intelligent circuit design in the location of the crystal oscillator. There are very extreme converters that do fancy things to improve clock jitter specs, but it's been proven that even those don't help anything with most setups when you must use one clock to control any external conversion to keep it in sync, because ANY converter that is externally clocked, even the very best, suffers greatly from being externally clocked.

Oh, and for the person who was saying he thinks the definition of an audio interface is a converter with pres, this is incorrect. an audio interface is one or more line amps/input buffers going into a/d converter(s) and one or more d/a converter(s) going into line amps/output buffers.

In other words, every single audio "converter" product that has audio connections is actually an audio interface, not just a converter. The converter is just a small part of every audio interface.

You can't technically buy a standalone converter since it won't do anything useful in your studio due to not having -10/+4 unbalanced/balanced ins and outs. -10 or +4 1/4" inputs and outputs require analog line amps and so on to get the signal levels right. The analog input to an a/d converter won't accept those levels directly, it requires analog circuitry.

Any audio interface that also has mic pres is an audio interface with mic pres. but EVERY converter product has line level preamps which are just the same thing as mic preamps but with lower gain and different impedance and other characteristics.

[smoovemode]

Ok, let's say I buy a Rosetta 200/800? Other than mic preamps and a firewire card,what would I need to get sound out of it?

[adrianww]

Originally Posted by mitchiemasha

Prefer Firewire or pci express? Not usb, My Virus TI runs on usb and sharing the ports can mess things up. PC, windows 7, 64 bit, running cubase 5. Min 6in 6out, seriously low latecy. All the other trimmings I have no need for. Need the AD conversion to be the best available. Not fussed on amount of Preamps. As for price, I just don't want to be paying more for features I don't need.

For those requirements, the first place I'd probably look would be RME Fireface 400. It's got enough analog I/O (2 pres, 2 line/instr inputs and 4 fixed level analog lines, plus 6 balanced analog outs and 2 unbalanced) along with SPDIF and ADAT options for digital expansion. It's a FireWire connection into your PC and RME have probably got one of the better reputations for driver performance/stability at low latencies. They also tend to have a reasonable rep for their conversion.

It's not going to be the cheapest option out there, but is probably well worth the money.

[drumdrum]

Its amazing the amount of people here on GS that bash the mr816 based from the "results" from an grossly inaccurate "test" that was done here.

unless you have done your own analysis of the device, please stop bashing genuinely superb equipment based on inaccurate tests.


The bottom line is this....

Stats matter for squat for 99% of users here on GS. Unless you have a great listening/mixing environment, you will never hear the subtle differences in todays generally high standard-to-low-cost ratio.
there are far more important factors with a mix than your converters harmonic distortion rating.

Pick an interface based on the I/O you need, preamp quality (if you aren't using external ones), cost, and your ability to work well with one. If a particular interface isn't doing it for you, for any reason, return it and try another.

Make music based on realistic requirements, and not so that you can feel like an audiophile and convince yourself you can hear the 0.001% difference...

Sent from my GT-I9000 using Gearslutz.com

[smoovemode]

Originally Posted by drumdrum

Its amazing the amount of people here on GS that bash the mr816 based from the "results" from an grossly inaccurate "test" that was done here.

unless you have done your own analysis of the device, please stop bashing genuinely superb equipment based on inaccurate tests.


The bottom line is this....

Stats matter for squat for 99% of users here on GS. Unless you have a great listening/mixing environment, you will never hear the subtle differences in todays generally high standard-to-low-cost ratio.
there are far more important factors with a mix than your converters harmonic distortion rating.

Pick an interface based on the I/O you need, preamp quality (if you aren't using external ones), cost, and your ability to work well with one. If a particular interface isn't doing it for you, for any reason, return it and try another.

Make music based on realistic requirements, and not so that you can feel like an audiophile and convince yourself you can hear the 0.001% difference...

Sent from my GT-I9000 using Gearslutz.com

I feel you