Topic: Edirol R-4 - Oade T-mod v. Stock: gain / self noise measurements

[Brian Skalinder]

I did this little test a while ago, but didn't get around to posting it.  I may or may not do the same tests for the new T-MOD v. W-mod I just received from the Oade Bros.  Anyway, FWIW...

TEST INFO

I started off with a simple test to identify the gain differences between the R4 stock and Oade Transparent MOD channels, since I noticed the T-MOD offered less gain than stock.  It eventually grew into the additional tests below, but here's how I did it:

• Generated a 1 kHz tone in Audacity at 24-bit / 96 kHz
• Looped playback, PC digi-out > Bel Canto DAC 1.1 > Segue Enterprises (Todd R) Dogstar RCA cables > RCA | 1/4" adapters > R4 (would've preferred XLR adapters, but didn't have them handy)
• Recorded mic-in with gain knobs all the way down
• Recorded mic-in with various "o'clock" gain levels
• Recorded line-in with gain knobs all the way down
• Recorded line-in with various "o'clock" gain levels

GAIN MAP

Turns out the Oade Transparent MOD (in its original incarnation...just got my updated T-MOD today) offers less gain than stock.  Here are the gain differences mic-in for the Oade T-MOD v. stock, feeding both the T-MOD and stock channels the same 1kHz tone signal:

==========================================
MIC-IN MIC-IN MIC-IN MIC-IN MIC-IN MIC-IN
------------------------------------------
    STOCK      |      T-MOD      |  GAIN
------------------------------------------
dBFS  |   +dB  |  dBFS  |   +dB  | O'Clock
==========================================
 -42  |     0  |   -30  |     0  |  07:00 \
 -40  |   + 2  |   -30  |     0  |  08:00 - I found 0 gain across this range strange, not sure how to account for it
 -38  |   + 4  |   -30  |     0  |  09:00 /
 -35  |   + 7  |   -28  |   + 2  |  10:00
 -30  |   +12  |   -26  |   + 4  |  11:00
 -25  |   +17  |   -22  |   + 8  |  12:00
 -21  |   +21  |   -20  |   +10  |  13:00
 -16  |   +26  |   -16  |   +14  |  14:00
 -11  |   +31  |   -11  |   +19  |  15:00
 -06  |   +36  |   -06  |   +24  |  16:00
   0  |   +42  |   -01  |   +29  |  17:00

And here are the gain differences line-in for the Oade T-MOD v. stock, feeding both the T-MOD and stock channels the same signal (but obviously a slightly stronger signal than in the mic-in test in order to achieve proper dBFS levels):

==========================================
LINE-IN LINE-IN LINE-IN LINE-IN LINE-IN
------------------------------------------
    STOCK      |      T-MOD      |  GAIN
------------------------------------------
dBFS  |   +dB  |  dBFS  |   +dB  | O'Clock
==========================================
 -40  |     0  |   -30  |     0  |  07:00 \
 -40  |     0  |   -30  |     0  |  08:00 - I found 0 gain across this range strange, not sure how to account for it
 -38  |   + 2  |   -30  |     0  |  09:00 /
 -36  |   + 4  |   -28  |   + 2  |  10:00
 -30  |   +10  |   -26  |   + 4  |  11:00
 -25  |   +15  |   -22  |   + 8  |  12:00
 -21  |   +19  |   -20  |   +10  |  13:00
 -16  |   +24  |   -16  |   +14  |  14:00
 -11  |   +29  |   -11  |   +19  |  15:00
 -05  |   +35  |   -06  |   +24  |  16:00
   0  |   +40  |   -01  |   +29  |  17:00

NOISE MEASUREMENTS

I also measured the frequency response for both T-MOD and stock channels mic-in with no gain, and when adding enough gain hit -12 dBFS (15:00 Oade T-MOD, 10:00 stock).  I then repeated the FR measurement for line-in with no gain, and when adding enough gain to hit -12 dBFS (15:00 Oade T-MOD, 10:00 stock), though of course the tone signal was stronger in order to achieve the proper levels using the same relative gain settings.  The results:

Line-in, no gain

CHAN  |  GAIN  |  COLOR
=======================
T-MOD | 07:00  |  GREEN
STOCK | 07:00  |  RED

http://img.photobucket.com/albums/v245/bskalinder/r4/line_no-gain_linear.jpg
http://img.photobucket.com/albums/v245/bskalinder/r4/line_no-gain_log.jpg

Line-in, with gain

CHAN  |  GAIN  |  COLOR
=======================
T-MOD | 15:00  |  GREEN
STOCK | 10:00  |  RED

http://img.photobucket.com/albums/v245/bskalinder/r4/line_gain_linear.jpg
http://img.photobucket.com/albums/v245/bskalinder/r4/line_gain_log.jpg

Mic-in, no gain

CHAN  |  GAIN  |  COLOR
=======================
T-MOD | 07:00  |  GREEN
STOCK | 07:00  |  RED

http://img.photobucket.com/albums/v245/bskalinder/r4/mic_no-gain_linear.jpg
http://img.photobucket.com/albums/v245/bskalinder/r4/mic_no-gain_log.jpg

Mic-in, with gain

CHAN  |  GAIN  |  COLOR
=======================
T-MOD | 15:00  |  GREEN
STOCK | 10:00  |  RED

http://img.photobucket.com/albums/v245/bskalinder/r4/mic_gain_linear.jpg
http://img.photobucket.com/albums/v245/bskalinder/r4/mic_gain_log.jpg

CONCLUSION

So what does all this mean?  Well, unless my testing methodology is fundamentally flawed (which is entirely possible), I think it means:

<01>  The Oade Transparent MOD offers less overall gain  than stock.

<02>  The Oade Transparent MOD reduces self-noise significantly v. stock.

<03>  If you can get away with running line-in, do, since it adds less self-noise than mic-in (which I wouldn't have expected, since I assume line-in is just mic-in with a pad, but...perhaps that assumption is incorrect).

<04>  I don't know how much of the self-noise in the results is due to my Bel Canto DAC 1.1, and how much due to the R-4.

[balou2]

Line-in, with gain

CHAN  |  GAIN  |  COLOR
=======================
T-MOD | 15:00  |  GREEN
STOCK | 10:00  |  RED

http://img.photobucket.com/albums/v245/bskalinder/r4/line_gain_linear.jpg
http://img.photobucket.com/albums/v245/bskalinder/r4/line_gain_log.jpg



<02>  The Oade Transparent MOD reduces self-noise significantly v. stock.

GREAT comp Brian.  The noise floor is indeed much lower with the T-mod, but how much difference does the mod make when the noise floor is so low to begin with?

[Busman Audio]

You also need to take into account the fact that lowering the gain lowers the noise floor by nature. Not that the noise floor isn't lowered also.
I also lower the noise floor significantly in my mod now. It took a while to develop that part of the mod but I now offer it on all new mods and anyone that has one already can send it to me to get that part done.

My noise floor mod also lowers the gain enough to use the mic in more often.

I also offer  a warm mod.

Thanks for doing these tests Brian.  If I can ever get some time to do intensive testing I will post my results for my mod also here.

[Brian Skalinder]

You also need to take into account the fact that lowering the gain lowers the noise floor by nature. Not that the noise floor isn't lowered also.

Yup.  That would explain why line-in has lower noise than mic-in.  Also why I checked noise at no gain (as a reference point) and then again with some gain.  Too much work to check noise at all gain settings, so I picked a level above which I don't usually go - 15:00 on the Oade mod channels - and stuck with it.

Chris - do you know if line-in = simply mic-in with a pad?  Just curious...

[Busman Audio]

Yes that is basically what it is because it still goes through all of the same circuit that the mic in does. It switches the path to go through different resistors but the all of the same op amps and capacitors are involved.

Great work on the tests. You really work hard for this community. 

And you make great bluegrass recordings

 

[Brian Skalinder]

Great work on the tests. You really work hard for this community. 

Well...truth be told, most of what I do is purely selfish - I want to know what my gear sounds like, how it tests, etc., for my own purposes.  I just figure if I have the results, I might as well share.

Gotta pick up a pair of RCA-f > XLR-m adapters for duplicating this test on the W and new T mod, since this round of mods disabled the 1/4" inputs.