Hi there,

I don't normally use bypass mode. However, I recently inherited a decent turntable, and I wanted to use bypass mode last night.

I turned off my sub at the source, so there's none of that to get in the way.

So I'm listening to some Steely Dan on vinyl, and kick it into bypass mode; there was effectively no change. Still in bypass mode, I go into the OSD and set main speakers to "large". Now my bass is present in my mains.

This morning, I did a little test. This was using my PC with an optical connection. Send a 70 Hz tone to the 990. Enter bypass mode. Set crossover point at 100 Hz. Switch between large and small in the OSD. Bass comes and goes in the mains. No sub attached.

So what's up? The "990 Digital Bass Management Matrix" indicates that bypass mode will set the mains to "Full Range" regardless of all other settings, and in all cases. I note that the "990 Digital Bass Management Matrix" has an overall header of "Analog Input Signals". As noted above, I've tried this with both a PC as input to the 990 (digital), as well as a turntable (analog).

Anyone else see this?

Bypass with a digital input is not an effective test of how bypass works with an analog input. There are two separate tables for Outlaw's "990 bass management matrix" - one for analog input signals and one for digital input signals .

With an analog input, bypass mode will yield a full range signal to the mains no matter how the speaker's bass management is configured (large or small). The bass management setting does determine whether or not a duplicate of that analog input is copied to the subwoofer - when the fronts are set to "small" and an analog input is set to "bypass" the 990 will send a summed copy of the full range inputs to the sub, whereas when the fronts are set to "large" and an analog input is set to "bypass" the 990 will not send anything to the sub.

Hi gonk,

I just did another test. First set my PC to output analog signals. I figured there was some difference in using digital signals. I used a Galaxy CM-140 SPL meter to take measurements. I had C-weighting enabled and "Slow" mode selected. I was in the "Medium" range (I think it's 50 to 100 dB).

First, crossover frequency did not seem to make a difference. Also, the sub selection of "LFE Only", "L+R+LFE", etc. did not seem to make a difference.

All measurements were taken using Bypass Mode.

So, to recap, I'm taking SPL measurement using pure tones sent using an analog signal from a PC. The 990 is in Bypass Mode for all measurements. The only thing I'm changing is the status of the mains from "Large" to "Small". I also scrolled through crossover settings and sub "configs" while measuring, and those seemed to have no effect.

Freq=76 Hz
Mains Lg = 73.1 dB
Mains Sm = 69.1 dB

Freq=50 Hz
Mains Lg = 63.4 dB
Mains Sm = 53.8 dB

I would expect there to be no change for the mains SPL readings based on Large or Small settings when using Bypass Mode.

Am I missing something?

Hmmm... I'm not sure on this one.

You've got an analog input in both cases, with the mode set to bypass in both cases, with the volume set to the same position in both cases, and the only difference was whether the fronts were set to "large" or "small"? The fact that changing the crossover point didn't affect the outputs suggests that there really is no crossover being applied, assuming that at least one crossover chosen was below one or both of these frequencies. I'd be curious to know what is happening when the frequency is something well above the bass management/subwoofer domain (maybe 500Hz or 1,000Hz), just to see if the ~4dB difference is happening across the board or if it is limited to the range in which a subwoofer would also contribute.

A little clarification of the thinking behind that last comment... The consistent 4dB difference makes me wonder if there was an attempt to compensate for the presence of a subwoofer to prevent double bass by dropping the level 4dB below a certain point. I hadn't ever heard of such a feature, but it would be a reasonable explanation (and a sort of cool feature, actually).

Hi again,

Yep, you've got it right. An analog signal coming from my PC playing a single tone in bypass mode. All I'm doing to alter things is switch between small and large size for my mains speakers in the OSD. Sub is unplugged and I'm not messing with the volume control.

Same test at 1kHz measured no change.

Freq=200 Hz
Lg=70.3 dB
Sm=69.8 dB

Freq=100 Hz
Lg=78.7 dB
Sm=77.3 dB

FWIW, the front crossover point was set at 40 Hz, but moving it around didn't appear to make any difference in the measurements.

I see what you're saying about dropping off 4 db for any potential sub contribution. However, since they're publishing that the mains be full range, I don't think there should be any frequency roll off at all.

Also, the sub output in bypass mode is also full range, so they couldn't assume at what frequencies the sub would contribute. I'm using a DIY sub that relies on BM from the 990, so full range into that sub sounds pretty bad. Or, what if I had an SVS or Outlaw sub and was using its built in crossover? They wouldn't know where that was set in order to roll off my mains.

All in all, I think that full range should be full range, no roll off, when in bypass mode, regardless of the size setting for the mains.

Finally, if they are rolling off the mains at whatever point, and for whatever reason, does that imply that they have sampled my analog signal?

It's probably in the manual somewhere, but do they directly state that bypass mode means no A/D and D/A?

Thanks!

It might be worthwhile to run the question past Outlaw's customer support. For what it's worth, here's the reasoning that I'd see behind a 4dB drop for frequencies below 100Hz or so when using analog bypass and small speakers.

First, the logic behind Outlaw's decision to send a copy of the signals to the sub when front speakers are set to "small." If you have the speakers set to "small" then presumably they are not good down to 20Hz - or for that matter, possibly not even 60Hz or higher. That's the reason for the 990 to still send a signal to the sub with small speakers when in stereo analog bypass in the first place: those speakers, being identified as "small," are going to naturally roll off somewhere below 100Hz but a fair bit above 20Hz. Sending a full-range signal to a sub isn't necessarily an ideal situation no matter whose sub you have, but if you are likely to use this arrangement with the 990 I would tend to suggest something that I otherwise recommend against with surround systems: enable the sub's internal crossover to prevent it from having to deal with that full range signal. Set it high enough to keep it from interfering in normal circumstances, but let it provide a "backstop" of sorts to high frequency material.

Now, we'll carry it one step further. Say that we're assuming that speakers are rolling off naturally because they are identified as "small" but we have no idea where that roll-off is happening. We do know that the sub is getting a copy of that data, and it's going to try to play at least everything below 80Hz (even with a sub's crossover engaged, this is a pretty safe bet). That means that we've got a "double bass" situation potentially happening. How do we minimize this? (Especially considering the endless debate that the idea of "double bass" generated for the old Model 950.) Well, we know that there's a sub in the system to help below 80Hz because the 990's set up that way. That means that we could reduce the gain a bit over a small portion of the low end so that the reinforcement provided by the sub won't be as overpowering. Because there are too many unspecified values, it's going to be a rough correction at best, but there's a certain curious logic hiding in there. Like I say, though, this is just me musing over random possibilities - I've not heard of this actually being the case.

I don't know where it is specifically listed, but yes - bypass definitely means no A/D/A. Anything happening to that signal is happening in the analog domain.

Hi gonk,

I talked with Steve at Outlaw, and he said it shouldn't be doing that at all. "Full Range" to mains shouldn't have any rolloff anywhere.

I agreed to do a system reset, reconfigure all my settings and see if the problem still exists.

After that, he suggested I try reinstalling the firmware. Depending on how I'm feeling tonight, I may try that as well.

If it persists after that, Steve's indicating to me that there's a problem.

So, gonk, if you're up for it tonight, see if you can recreate the problem. It's simple and you don't really need to measure. Just play an analog source in bypass mode. Switch mains between large and small. You should hear no difference. Although I measured it to obtain a little more detail on what's happening, it really wasn't necessary; it's rather easy to *hear* the difference.

I'll repost after the system reset.

Thanks!

Two other things:

1) The title of the thread is a bit misleading, as I don't think that bypass mode is really applying bass management. It may be using pieces of the algorithm somehow, but it's not as simple as saying BM is still working in bypass mode, as crossover point and other settings are still irrelevant.

2) Although bypass mode should mean no A/D/A, I'm very curious to know in what domain my rolloff issues are occurring. I would not be surprised if it were happening digitally, as I wouldnt' expect the 990 to have filtering implemented in analog circuitry (I could be wrong, but I don't see a need for analog filtering).

The same issue occurs after a system reset.

Half way through setting *everything* up in the OSD, I realized that maybe I shouldn't have put so much time into it if I'm just going to do a FW flash, 'cause I'll just have to re-enter everything after the FW flash. Oh well. Maybe later tonight.