Topic: M10 vs 722 Comparison files & poll

[johnw]

Material is phish Alpine Valley 8/15/2010 Oh Sweeth Nothin
Sonosax gain set at roughly 10:00 both channels on low gain

Schoeps MK41>Nbox>Sonosax SX-M2 xlr out>Sound Devices 722 (+1.8 gain)>wav(24/48)
0.301 dB gain added in wavelab to bring peak level to 0

Schoeps MK41>Nbox>Sonosax SX-M2 1/8" out out>Sony M10 (gain dial set at 4)>wav(24/48)
no gain added as peak level was 0

Download files here:
A: http://www.megaupload.com/?d=VUOZ0LPV
B: http://www.megaupload.com/?d=WG0C4CM0
md5:
63f7cdb03e1e305b359952023a7e5bf6 *osn.A.flac
42f13d9c86e69998c413fa219a3cb4e7 *osn.B.flac

[johnw]

Also, I realize that the right channel is about 2dB lower in both sources. Not sure of the reason - normally with the Nbox and the matched MK41 caps the channels come out even on the 722 with linked gain. Don't think that affects the test at all, but feel free to correct it.

[it-goes-to-eleven]

Thanks!  D/ling now...

Generally, I do think it is essential to match the db's in a comp.  Tests have shown we are greatly biased to pick a louder source.

[page]

That's tough. Starting at about 1:50 for about 15 seconds was one reference point, another was the first 6 seconds, and a third was starting at 2:45 for about 7 seconds (how does it handle what appears to be analog distortion). I think I prefer A, it seems to have a little more shimmer in the low piano notes and the drums sound clear/intimate. The other one maybe has a mildly deeper fluffy bass?

This is all splitting hairs though, as much as the McSpendy box sitting next to me would like a clear winner, I'm not sure I can say there is one.

[johnw]

Thanks!  D/ling now...

Generally, I do think it is essential to match the db's in a comp.  Tests have shown we are greatly biased to pick a louder source.

I think the 2 sources are close to being equally loud. B might be a dB louder. I corrected my earlier post to say that the right channel was quieter in both sources. Or were you referring to something else?

[page]

Btw, if anyone else finds an earth shattering passage that I missed, I'd like to know so I can go back and listen again.

[johnw]

I figured 24 bit would be the best choice for most people. Should I dither these down and post 16bit?

[page]

I figured 24 bit would be the best choice for most people. Should I dither these down and post 16bit?

I'd think if we are comparing A/D stages, as close to the original source as possible is preferred.

[it-goes-to-eleven]

I think the 2 sources are close to being equally loud. B might be a dB louder. I corrected my earlier post to say that the right channel was quieter in both sources. Or were you referring to something else?

I misread your post, sorry.   The average rms levels between sources are extremely close, so no worries.

[notlance]

I could not tell enough difference to be able to vote, much less have an opinion to which was better.  Any vote I'd make would be a WAG.  However, I expected to hear little or no difference so I am sure that biased my results.  Maybe I'll listen again later to hear if I can tell a difference.

Perhaps with a recording with more quiet passages it would be easier for me to tell them apart?

[page]

I could not tell enough difference to be able to vote, much less have an opinion to which was better.  Any vote I'd make would be a WAG.  However, I expected to hear little or no difference so I am sure that biased my results.  Maybe I'll listen again later to hear if I can tell a difference.

Perhaps with a recording with more quiet passages it would be easier for me to tell them apart?

Yeah, in that regard, I'd be tempted to find some stiff attenuators to drop the signal almost 30db or so and then record it to both. Figuring that would require amplification later and reveal some of the lower level flaws of the A/D stages. However, getting to that sort of points out for, that what many people here use these boxes for, it's a tough call if you are using the SD boxes as A/D stages only. Now, if you are using other SD features, then there is a better price justification (redundancy in recording or powering being prime examples, P48 might be another).

In an armchair type way, I'd be interested to see what the folks over on GS have to say about it.

[Nick's Picks]

downloading now...will listen, vote and respond shortly

[notlance]

Perhaps a vocal music recorded in a large quiet church where you can hear the music decay to the background noise level, and recorded at -40 dBFS would be a good test.

The point is not that I can conceive of a test that might show the audible difference between the A/D section of a $2000 and $200 recorder.  Rather, I think it is more significant that an extraordinary test is required.  There are many good inexpensive (and expensive) recorders out there.  Maybe there are more important aspects of recording to focus on.  I like gear as much as the next guy, but so often we are too focused on the "noise".  The problem with my recording are that I placed the mics too far away, or maybe the flute player should be turned 45 degrees, or the HVAC system is way too loud, or I'm picking up RF, or ....   Never have I thought "this recording would be perfect if only my A/D was better".

[johnw]

If another test needs to be done, I'm going to have to borrow another preamp. The other thing I hoped to accomplish with this test was to establish what unity is on the M10. Without giving anything away, I'm pretty confident it isn't 5 and I don't think it is 4. To establish that though I need to know what the difference (if any) between the 1/8" and xlr out is on the Sonosax SX-M2 and unfortunately nobody has answered that in the other thread. Since I don't record vocal music in churches, what this test does answer is whether I absolutely need to bring the 722 to get a listenable recording and for me that answer is no.

[Nick's Picks]

again...we kill our own logic by disputing such things that are totally immeasurable in an environment such as a phish concert.
that does jab at a lot of on this list....the concert tapers, IMO...being so fickle about these things that can not be measured consistently or reproduced for that matter.

pointing the finger directly at myself with that comment..., btw.