[Flea Bag]

Az B,

Thanks for the manufacturer's insight. I've discussed the issue of output 'clipping' with someone last week and felt that I need to clarify or better phrase my wording. The 'clipping' I was originally referring to was the User Manual's definition of clipping which is actually just a form of frequency limiting which cuts only the particular frequency which exceeds the equalizer's output capability. 'Frequency chopping' will be what I'll call it from now onwards or if someone could name the proper technical term.

With regards to the more common definition of clipping, like jhenderson010759 mentioned, the equilizer does encounter very noticable clipping across the whole frequency spectrum if the frequency chopping reaches more servere levels. My frequency chopping only involves amounts of less than 0.7dB -too small for me to hear any difference. If one were to equalize more serverely than I do, the DEQ2496 will indeed exhibit clipping as opposed to just frequency chopping. This is most obvious when playing around with the DEQ function of the 2496.

As for good recordings, they never seem to exceed the -0.0dB peak reading. However this does happen to some very well known various-artists-compilations, hastingly put together. This is especially so with a particular pirated CD I tested. Now we have aural and measured proof of bad recordings!

So for new 2496 owners; If your optical digital input shows 'clip' as a peak reading, its more likely that your recording is bad than your equipment is faulty. However, if you still get 'clip' as a frequent reading even with well mastered CDs, then your DEQ or source equipment could have a problem.

I forgot to mention that if you are boosting any particular frequency until the out level indicates 'clip', you can:

a) Reduce gain offset in page one of the utility menu.(as stated in my previous post)

b) Do it directly from the GEQ page. (THIS IS NOT LISTED IN THE INSTRUCTION MANUAL!!!) Press the upper (small) data wheel to change the FREQ selection to read 'FULL'. Now you can use the large data wheel to horizontally offset the entire frequency range.

c) If you would like a bass boost for example any range below the 80Hz mark, the PEQ can also serve as an alternate method of preventing output clipping/frequency chopping. In the PEQ page, you could turn the lower small data wheel until it reads 'L6dB', 'L12dB', 'H6dB', 'H12dB', 'LC' or 'HC' and then make the bass more prominent by reducing all other frequencies.

On the topic of bass boosting, I've got a quick and convenient tip to make the DEQ2496 operate with the convenience of a 'Mega Bass' boost button you see on a portable Discman. I save most of my settings for my HD280 in GEQ mode, keeping them neutral.(still haven't found what I call neutral yet!)

Next, I make a bass boost in PEQ mode and save both PEQ and DEQ as a single preset which I call preset A. So preset A allows me to listen to my music in a nice neutral tone (or so I would think) and when I'm not happy with the bass, I simply press and hold the PEQ button to activate/deactivate the bass boost. No need to switch between presets.

That's all I have time for at the moment!

gerG, do you think it would be good if you put up some kind of review/FAQ for the DEQ2496? It's obviously quite popular a piece of equipment around here and the instruction manual alone just isn't enough. Perhaps it'll be more of an extended instruction manual than just a review. If you would like me to put all my findings about 'clipping' in one post for the review, just let me know.

 

[tomek]

i'm all for a FAQ!

the tip i got to lower the delta between adjacent frequencies when using auto EQ really helped. it was a big difference and nothing that i found in that skimpy user's manual.

 

[Az B]

Quote:

Originally Posted by jhenderson010759 Az B - What are you using as a digital volume control?

I have a Carver Reference CD player with a volume control.

Az

 

[jhenderson010759]

Quote:

Originally Posted by Az B ... Originally I was not planning on inserting <the 2496> into the loop, but since I can use the digital in, and then the digital out to the DCX, and the DCX has limited processor power for EQ, I have it in the loop and it works great.

What data format do you use for the digital connection between the DEQ and DCX? I need to use SP/DIF from my CD transport to the DEQ, but it isn't clear from the DEQ manual whether it will simultaneously drive the AES/EBU outputs when receiving SP/DIF inputs via TOSLink. Is that your configuration also?

 

[gerG]

Jim, sweet speaker project! I use a toslink into the DEQ, and AES balanced outgoing. That line at the bottom of my sig is pretty much the settup. For speakers, I have been using the dbx digital xover with high pass directly to Morel tweeters, with never a blip. I do the same trick in my car, with a dedicated amp just for the tweeters and an active crossover. I was worried about a stray chirp taking out a tweeter, but it has not happened in 2 years of operation.

I am getting very curious about the DCX. I love the idea of going digital straight into the x-over. I am suspicious of digital volume control, though. Does that not reduce amplitude resolution? I am thinking analog volume controls downstream of the D/A step. As touchy as my amps are, I don't need any extra gain in the loop.

Flea, I think that a FAQ is an excellent suggestion. This thing has started to catch on lately, and the info should be easier to mine than reading through this whole thread (and the twin thread that is burried at the moment). Thanks for mentioning the trick for changing volume in the GEQ window. That feature is actually unique to the latest firmware update, I think. I always bounce back to the meter window to check my levels (just like the old tape days). When I am finished with adjustments I leave the display in the rta mode.

Az B, welcome aboard! Have you tried the DEQ with headphones yet? You are right about the resolution of the built in RTA. For people who have a test mic already I recommend picking up a decent PC interface/preamp and a copy of TrueRTA. Take a peek at my room acoustics thread for an example of what it can do. For a $100 software package it is just short of miraculous.

A parametric eq can help with the LF modes in a room, but it is better to get at the source if you can. Pulling down a reinforcement with a parametric resolves the quantity issue, but now you have a signal that is roughly half source and half 360 deg shifted reflection. The balance will be right but you are giving up transient response. You will still be left with the cancellations.


gerG

 

[gerG]

Forgot to mention another subwoofer on my must build list. The Peerless XLS 12" with the 12" PR is an amazing combination in a small cabinet. Peerless has an excellent white paper on their website:

Peerless 12

Don't be put off by the disclaimer, it is a good read.


gerG

 

[ooheadsoo]

Oh yeah, the peerless is supposed to be sooooooooome really really good driver, maybe one of the best. I had been considering that driver until this one popped up: http://www.gr-research.com/sub.htm

According to my sims, I can get -3db at 22-24hz or so with the stock PR.

Sound should be decent, I would think. The only thing I know gerG doesn't like is the foam surround, but I have no problems with them

 

[jhenderson010759]

Quote:

Originally Posted by gerG ... I am getting very curious about the DCX. I love the idea of going digital straight into the x-over. I am suspicious of digital volume control, though. Does that not reduce amplitude resolution? I am thinking analog volume controls downstream of the D/A step. As touchy as my amps are, I don't need any extra gain in the loop. ...For people who have a test mic already I recommend picking up a decent PC interface/preamp and a copy of TrueRTA. Take a peek at my room acoustics thread for an example of what it can do. For a $100 software package it is just short of miraculous. gerG

gerG -

I agree that a digital volume would reduce the amplitude of the outbound signal, signal purity should not be affected. Digital volume should be implemented as a scalar multiply, which is a linear operation. True, the dynamic range of the output will be reduced (that's the whole point), but the frequency content will be unaltered. And I'm only interested in unity gain or less (attenuation operation only), which avoids saturation scenarios.

If I could locate a digital preamp featuring multiple analog and digital input sources plus a digital output amplitude which tracks it's volume control, then the DEQ and DCX could both lie downstream from this switch. All source matter could readily benefit, rather than just a single source, such as a CD, as in my current hookup. Basically, I want Pre-Out, which traditionally is volume controlled, in digital form. Unfortunately, none of the pre-amps or AV receivers I've seen do this.

What do you recommend for an inexpensive mic pre-amp? I have TrueRTA and the Behringer ECM8000 mic, and I'm thinking about either the Behringer Shark DSP 101 or the Behringer UltraVoice Digital Vx2496. The former is cheaper, but the latter has digital out, so I could skip a D/A -> A/D conversion on the way to my EMU 0404 sound card if I get a pre featuring digital out.

 

[gerG]

Good morning Jim.

I am using an M-Audio Duo usb interface right now. I would prefer firewire, but my Sony laptop has a rather quirky implementation of firewire, and it only works sometimes. usb is noisier, but it is reliable on that PC.

There are lots of options out there for a single box to do the preamp and A/D. I had a Behringer Shark, but it was just another piece of claptrap to wire up (and another wallwart). There may be units available that can be powered off the usb port. The firewire box that I had did that trick off my powerbook (true 6-wire firewire) so no extra wallwart/wire. Unfortunately TrueRTA only runs on a PC.

Another thought, does the Emu 0404 have phantom power and a mic input? IF so all you need is a long mic cable.

Let me know what you figure out on the digital preamp. I am still concerned that for large attenuation I will be losing resolution. I need to think about the math on that a bit. For small changes, not an issue.

gerG

 

[jhenderson010759]

Quote:

Originally Posted by gerG ...Let me know what you figure out on the digital preamp. I am still concerned that for large attenuation I will be losing resolution. I need to think about the math on that a bit. For small changes, not an issue. gerG

I think you're right about the resolution proportional to volume. I don't think distortion would vary, but that's not the whole picture. And, I certainly expect that I would be using substantial attenuation most of the time, so if it is a problem, it's probably not one that should not be ignored.

A quality, six-channel stepped attenuator between the output of the DCX and the amp inputs would do the trick, albeit in the analog domain. Do you know of any sources for one of these? XLR I/O would be nice.

 

[ooheadsoo]

How does this digital preamp stuff work? As far as i can tell, wouldn't you need at least 5 bits of volume attenuation data to get 31/32 steps of definition? That's a lot of data that would get truncated from a 16 or even 24 bit signal.

It would seem to me that the cheapest way (perhaps only temporary in the long run) would be to find a nice 6.1 receiver and instead of using the receiver for surround duty, use the discrete inputs to send the signal signal to each driver. That way you get unified volume control. I haven't heard of a 6 channel stepped attenuator

That would be monstrous! Maybe 10 inches long!

 

[jhenderson010759]

Quote:

Originally Posted by ooheadsoo How does this digital preamp stuff work? As far as i can tell, wouldn't you need at least 5 bits of volume attenuation data to get 31/32 steps of definition? That's a lot of data that would get truncated from a 16 or even 24 bit signal. It would seem to me that the cheapest way (perhaps only temporary in the long run) would be to find a nice 6.1 receiver and instead of using the receiver for surround duty, use the discrete inputs to send the signal signal to each driver. That way you get unified volume control. I haven't heard of a 6 channel stepped attenuator That would be monstrous! Maybe 10 inches long!

ooheadsoo -

DACT makes a six channel attenuator, but it's about $600. Atop that, I'd have to buy the case, connectors and build it. MSB makes an active attenuator, for about $800. I can't believe that I need to spend that much money for a volume control. There's got to be a better way.

Regarding the AV receiver, that is my current setup. But, it suffers from this problem, and is under-powered for the subs. Optimally, I'd like to be able to use a mix of amps as required for the different drivers in the system.

 

[gerG]

Jim, there are similar units to the DACT, but none are cheap. There is a stepped attenuator thread over in the diy section. I bet that company can stack wafers. That would end up at $180 for 6 channels. Not bad for attenuators of that quality. I actually like the idea of a long narrow housing with the inputs on one side and the outputs on the other. you could use XLR, but forget going fully balanced on the attenuators. You would need 12 rotary switches for that trick!

Another approach that I am considering is to build a 3 position volume control, then use the gain control within the crossover to fine tune. I am assuming, of course, that there is a global gain option in there.

You could also use integrated amps and use the built in volume controls to set your baseline, then tweak down with one of the digital boxes.

I haven't looked, but there are bound to be multi-channel passive preamps out there. In fact, there must be used stuff in circulation.

I will look for other options this weekend. Even though my current setup is analog in, I have the same problem because the amps are so touchy that the x-over is operating down in the weeds. I thought that it had a gain offset option, but that seems to be on the input, with helps nothing. Resistors or transformers on the output are my only options, other than switching back to an analog crossover (blech). I suppose I could listen with earplugs in and turn loose 2 kilowatts, but I don't feel like repairing water lines again.


gerG

 

[jhenderson010759]

gerG -

I found a Roland M1000, 24-bit digital mixer for $299 on the web and ordered one for experimentation. This unit can perform the digital attenuator function, operates with 24-bit resolution and allows sourcing my CD player and/or my PC via USB to the Behringer DEQ and DCX. I don't think resolution loss will be an issue. Here's why:

If I upsample from 16 to 24-bits before sending to the Roland, I should be free to effect an 8-bit attenuation, approximately 45 dB, with zero theoretical resolution loss. This is because upsampling produces samples with valid information in bits 23..8 and either dither or zeros in the least-significant byte. So, when the Roland attenuates the input signal using 24-bit arithmetic, the least-significant eight bits of each sample may be truncated without a loss in resolution.

As my entire CD collection is stored on my PC in .ape format, I can use Foobar to upsample to 24-bits during playback - just as I do in my headphone setup. If I can find a a CD transport that up-samples, I should be in good shape.

What do you think?

 

[gerG]

That sounds like a great approach. It was also quite informative. Is this a common approach for upsampling? I had always assumed that they interpolated to get to higher resolution, but this is certainly easier. I should think that a 45 db envelope would be plenty. Plus you get to have yet another digital toy in the loop!

As an alternative, you could also do the digital volume control on the PC after upsampling.


gerG