Topic: R09 R09HR & R44 internal mic preamp comparison (indoors test).

[digifish_music]

Aim:

To compare the internal microphone preamplifiers of the Edirol R09, R09HR and R44 digital recorders using a pair of low noise stereo microphones (Audio Technica AT3032), recording in a quiet environment (indoors) so background noise levels of the systems can be easily compared.   

Methods:

Using the following...



Notes: The ART Phantom II (shown above) was used to provide Phantom power to the AT3032 microphones as the R09 and R09HR do not have 48V phantom power. This power-supply (2 x 9V batteries) does
not add any significant noise to the throughput signal. The ART Phantom II was also used with the R44, rather than use the R44's phantom power, to make comparisons and signal chains as equivalent as possible. However, rather than the adapter cable shown connected to the R09HR in the shot above, the short black XLR connections were used. This means the R44 was receiving a balanced signal while the R09 and R09HR
were not.

1. Levels

I attempted to match all recorders so that the peak-meters hit the same max levels = -12 dB. The recorder settings used were:

AT3032 Mics --> ART Phantom II --> R44 = -50 dB (2nd highest gain)
AT3032 Mics --> ART Phantom II --> R09HR = High-gain @ 50/80 (62%)
AT3032 Mics --> ART Phantom II --> R09 = High-gain @ 20/30 (66%)


2. Other settings

All recorders: Low cut off, plugin power off, screens minimum brightness, 44.1 kHz @ 24 bit.

3. Source

I recorded a test file of layered Sampled Piano and Synthesiser notes on my lounge HiFi

Here's what the perfect recorder (and microphones) would look like as this is the source audio file spectrograph...in particular look at the background, it's black, this means there is no noise. Recorded information shows as colored bands, that's OK, you want that



download the original here (2.8 Meg wav file)

Results:

Here is the spectrographs of all the recordings. In particular look at the decreasing contrast between the background and the recorded sounds as you move from top (best) to bottom (worst) below...



You can download them here (wav and mp3 versions)...

R44

R44 Recording (2.8Meg wav)
R44 Recording (740k mp3)


R09HR

R09HR Recording (2.8Meg wav)
R09HR Recording (740k mp3)


R09

R09 Recording (2.8Meg wav)
R09 Recording (740k mp3)

More detail?

In this recording I located the mechanical sounds of the CD player starting up at the beginning of the recording...



Download the CD Player noise here (800 k wav file).

Conclusions:

I am surprised, the internal mic-pre on the R09HR is, from an audible noise perspective, very similar to the R44 (the R09HR perhaps being just a tiny bit worse, or it could be level differences), the R09 is a distant third.

The R44 does perform better than the R09 or R09HR in terms of digital noise perspective (follow this link to see the R44 vs R09HR digi-noise), but that is not audible.

Given that this is the TapersSection and not 'Bird-Stalkers-Anonymous', all three would be very good for recording live bands. The R09HR and R44 would even be good for unplugged sessions

I will post up the MixPre --> Line in vs the above next.

Enjoy!

The internal microphones for these recorders are compared here...

http://taperssection.com/index.php/topic,108779.msg1452098.html#msg1452098

digifish

[Eigenklang]

Great work!

The 09's internal amps are just not the way to go even with background noise. All in all, recording less-than-loud things with the R09 can be quite expensive 

[digifish_music]

Great work!

The 09's internal amps are just not the way to go even with background noise. All in all, recording less-than-loud things with the R09 can be quite expensive 

Thanks

I've done quite a few binaural field recordings with the R09, I have been happy with the results. For example, most of the stuff here...

http://www.freesound.org/packsViewSingle.php?id=2452

That said, I am even happier with the HR

digifish

[Carlos E. Martinez]

Conclusions:

I am surprised, the internal mic-pre on the R09HR is, from an audible noise perspective, very similar to the R44 (the R09HR perhaps being just a tiny bit worse, or it could be level differences), the R09 is a distant third.

The R44 does perform better than the R09 or R09HR in terms of digital noise perspective (follow this link to see the R44 vs R09HR digi-noise), but that is not audible.

Given that this is the TapersSection and not Bird-stalkers-anonymous, all three would be very good for recording live bands. The R09HR and R44 would even be good for unplugged sessions

My guess is that both Edirol (R44 and R09HR) might be very similar at the input, perhaps even using the same chip, implementing one in balanced mode. It would be nice to find a circuit diagram of both to check on that.

Perhaps Oade or others that modify these units can help us on understanding this.

[flintstone]

Great comparison!  I guess it makes sense that the preamps of the R-44 and R-09HR
perform the same since the products were developed at the same time, and the R-09HR's
pre is touted as a big improvement. (Edirol calls it the Isolated Adaptive Recording Circuit).

The pricing of the two recorders also supports the idea that similar electronics
are inside. Edirol's retail price of the R-09HR (two channels) is about half that of the
R-44 (four channels).  Digifish's previous reports show that Edirol did save money on
the mics built into the R-44.  They perform poorly compared to the R-09HR's mics.

Now let's encourage retailers to discount the R-44 the way they have lowered the
price of the R-09HR!  The best I could find is $795 at Audiomidi.com.

Your previous results showed that the R-09HR's performance was improved by
using the Sound Devices MixPre instead of the internal pres.  So it's a fair guess
that the R-44's sound would be improved by using a four channel mixer like the Sound
Devices 442 ($2500) or even two MixPres ($1330 for two).

I know that managing three separate devices (two MixPres and one R-44) would be
a huge pain for field recording.  If you could fit it all into a gear bag, I think the
results would be fantastic, at a very affordable price for a quality four channel recorder.
($795 for the R-44 + $1330 for the MixPres = $2125).

If I remember correctly, Oade's R-4 Pro concert upgrade also costs $2000.  Anybody
up for a (R-44 + MixPre) versus (Oade R-4 Pro upgrade) comparison?

Flintstone

[digifish_music]

Great comparison!

I know that managing three separate devices (two MixPres and one R-44) would be
a huge pain for field recording.  If you could fit it all into a gear bag, I think the
results would be fantastic, at a very affordable price for a quality four channel recorder.
($795 for the R-44 + $1330 for the MixPres = $2125).

...

Flintstone

Thanks.

While it would be a bit clunky (but possibly not so bad in the right bag), a R44 and a pair of MixPres would be hard to beat for 4 channel pristine ultra low-noise recording.

digifish

[Ozpeter]

I don't think we've yet fully explored the R-44 noise levels other than at max gain have we?  In other words, recording a clock at 24 bits at (say) the noon level of the outer ring, then normalising, vs recording the same at full gain and normalising that?  I know that when the Zoom H2 was tested that way, it was evident that the "high" gain setting had the same signal to noise level as the medium gain setting, and therefore there was no point in using it - safer to record on the medium gain setting and normalise later.

[stevetoney]

Can someone help me please?  You've reached the conclusion that the R-09 is clear third.  What do you see in looking at the spectographs that eable you to reach that conclusion?  I don't understand because they all look almost the same to me, so I'm no debating...just trying to understand how to read the data that you've shown.

Thanks!

[digifish_music]

Can someone help me please?  You've reached the conclusion that the R-09 is clear third.  What do you see in looking at the spectographs that eable you to reach that conclusion?  I don't understand because they all look almost the same to me, so I'm no debating...just trying to understand how to read the data that you've shown.

Thanks!

The spectrograph is just a visual confirmation (note the increasing fuzziness of the background, that's white noise. Perfect would be pitch black, as in the background of this one --> click here )



The real test is what you can hear...Download here

Download that, put headphones on, turn up the volume and listen. The R09 is not a disaster or unusable, I have been making good recordings with the internal pres for a couple of years, it's just not as good as the HR/R44. For most people here it's also less of an issue as you record loud bands in noisy environments.

PS: Apologies for making assumption, I have edited the first post to make things clearer.

digifish

[stevetoney]

Thanks for the explanation digi.  Good stuff! 

[digifish_music]

I don't think we've yet fully explored the R-44 noise levels other than at max gain have we?  In other words, recording a clock at 24 bits at (say) the noon level of the outer ring, then normalising, vs recording the same at full gain and normalising that?  I know that when the Zoom H2 was tested that way, it was evident that the "high" gain setting had the same signal to noise level as the medium gain setting, and therefore there was no point in using it - safer to record on the medium gain setting and normalise later.

Sounds like something worth doing...if I don't see you do it first

digifish

[Kindguy]

Good stuff. Thanks for the effort.