JubScala?

[Edgar]

If you're going to the trouble of doing a JubScala conversion (or buying Jubilees), you're looking for high quality sound, so there's no point in cheaping out on a crucial part of the signal path.

Amen. It amazes me that people will think nothing of spending thousands on a preamp, and then cheap-out on a $200 crossover. If you think about it, the crossover has a more difficult job than the preamp -- the preamp has to act as much like a "wire" as possible, and still sound good, while the crossover has to do some serious modification to the signal, and still sound good.

Greg

[Chris A]

I've not messed with the newest Ashly units, but the Yamaha SP2060 is a good alternative to the Dx38...

Mike,

Did you hear any differences when we heard the Yamaha unit last year vs. the Dx38? I remember the difference between the JM Peach and the other preamp we heard.

..."much more elegant signal path"...

Interesting.

Chris

[DrWho]

Ya, I was even able to blindly identify the Yammie since it sounds just like the new high end digital mixing console from Yamaha....and that's not surprising since it's the same signal path.

I don't think we had the gain structure set to get the Dx38 into the area where it swaps between A/D's since I wasn't able to hear the huffing sound you get when the noise floor dances around. There's also a graininess injected from the small non-linearity as the gain jumps slightly. It'll sound like a crossover distortion, but it's happening at a positive and negative voltage instead of at zero volts.

For what it's worth, these effects are small in the grand scheme of things, and most people at work don't hear anything in tbe blind tests until they've been taught what it sounds like. The distortion is also dependant on the parts tolerances (EV used 0.1% resistors), and a lot of source material should mask the noise floor pumping. Ironically, it's worst when maximizing the input signal strength into the EV....

[Chris A]

I don't think we had the gain structure set to get the Dx38 into the area where it swaps between A/D's since I wasn't able to hear the huffing sound you get when the noise floor dances around.

Ironically, it's worst when maximizing the input signal strength into the EV...

So--you're saying, don't use the Dx38 at high input gain and you won't hear "noise floor huffing". Do you remember the relative input setting of the Dx38 at that time. I don't remember the exact Adcom preamp model number during that time.

It'll sound like a crossover distortion, but it's happening at a positive and negative voltage instead of at zero volts.

You can hear noise floor changes at only the signal peaks? Interesting.

Chris

[DrWho]

The best way to hear the huffing is to play a low frequency sine wave and gradually turn it up....you'll need a preamp that can drive the input of the EV into clipping to guarantee you can recreate it. Also be sure to turn down the amplifiers so that the speakers aren't ripping your head off at full tilt. I think if you can drive 0dBV with the input gain on the Dx38 turned up all the way, then you should be able to get into the right region.

I'm gonna pull numbers out of the air, but what you have are two A/D converters with 90dB of dynamic range. One converter is used to cover 0V to say 0dBV, and the other covers 0dBV to +20dBV.

When the level gets above 0dBV, the first converter clips so the DSP switches over to the second converter that can go up to 20dBV. What's happening is that there is an amplifier that reduces the gain by 20dB before the second converter. Inside the DSP, the signal is then digitally amplified by 20dB so that the levels are at the correct amplitude.

At the end of the day, it'll look like you have 110dB of dynamic range, but in reality you're never hearing a signal with better than 90dB.

There are a million different ways to implement the gain ranging algorithm, each with slightly different artifacts, but at the end of the day your noise floor is always going to make a big jump when the signal crosses the threshold. And this is assuming perfect gain ranging.

In the real world, the amplifier before the second converter won't have a perfect -20dB gain, which means there will be a distortion whenever signals cross that amplitude. If it has a gain of -20.1dB, then the output will be 0.1dB compressed above that point. 0.1dB doesn't sound like a lot on the surface, but it's actually quite noticeable when it's a hard knee (the predominant distortions will show up as intermods).

If you're familiar with the companding artifacts of analog wireless microphones, then it's the same effect.

Putting it back in perspective, the distortion of your speakers will dominate the distortion from the Dx38, but it's just one of those things that is accumulative and ultimately not necessary with more modern units.

[Edgar]

When the level gets above 0dBV, the first converter clips so the DSP switches over to the second converter that can go up to 20dBV. What's happening is that there is an amplifier that reduces the gain by 20dB before the second converter. Inside the DSP, the signal is then digitally amplified by 20dB so that the levels are at the correct amplitude.

Sheesh, I'd have hoped that they would have gotten past that tired old design by now.

Way back in 1995 or thereabouts, when I worked in the Merlin group (now long gone) at EV, we did comparative listening tests between our single-range ADC and the then flagship Klark-Teknik (also part of EV) dual-range ADC that they were very proud of. The range-switching in the KT converter was brutally obvious to anyone who knew what to listen for (DrWho, you describe it perfectly), and we told them so. But I guess you can't teach an old dog new tricks. Here it is, fifteen years later, and evidently their dual-range ADC still has the same problems.

(Don't get me wrong -- it is still a very, very good sounding converter. It's just that more modern designs sound even better.)

Greg

[mikebse2a3]

never mind.

[DrWho]

Sheesh, I'd have hoped that they would have gotten past that tired old design by now.

Well the technique still works really well for oscilliscopes and other bench gear []

I don't think EV has touched the Dx38 since they came out with it in the late 90's. It's still using that ancient Motorola DSP too. However, I'd be surprised if their new DC-One had the same architecture. I need to get my hands on its service manual since it is cheaper than the Yammie unit.

You got any other xovers you'd recommend, Greg?

[curt248]

Going to buy 4 more la scalas tomorrow. That will make 9 total. Nine point one surround anyone? It's hard to pass up good deals on these speakers.

[Islander]

In the real world, the amplifier before the second converter won't have a perfect -20dB gain, which means there will be a distortion whenever signals cross that amplitude. If it has a gain of -20.1dB, then the output will be 0.1dB compressed above that point. 0.1dB doesn't sound like a lot on the surface, but it's actually quite noticeable when it's a hard knee (the predominant distortions will show up as intermods).

If you're familiar with the companding artifacts of analog wireless microphones, then it's the same effect.

Putting it back in perspective, the distortion of your speakers will dominate the distortion from the Dx38, but it's just one of those things that is accumulative and ultimately not necessary with more modern units.

If I'm reading you correctly, it seems like these artifacts would not be audible at typical home listening levels, and are only apparent at pro sound levels if you know what to listen for. I can live with that. No system is perfect, not even the extremely expensive ones, as you can read any given month in The Absolute Sound...

[Edgar]

I don't think EV has touched the Dx38 since they came out with it in the late 90's.

Holy cow, has it actually been that long? The older I get the more quickly time passes.

You got any other xovers you'd recommend, Greg?

Nope; sorry, I've been out of that end of the biz for quite a long time.

Greg

[Chris A]

If I'm reading you correctly, it seems like these artifacts would not be audible at typical home listening levels, and are only apparent at pro sound levels if you know what to listen for. I can live with that.

(With pre-apologies to Mike B. if I appear to be unrepentant on this subject):

I will here note that I've never seen noise figures on turntable/record inputs that exceeded 70 dB(A) of noise floor. The inherent EV Dx38 figures that are being bandied about here (90 dB- what scale?) are at least 20 dB better than records and significantly better than analog reel-to-reel tape (w/o noise reduction techniques). Why wouldn't we be talking about vinyl/turntable noise figures here for those folks that still use those devices on a regular basis. I'd hazard a guess that one of the least objectionable forms of distortion is noise floor--if it is very low as it is here--even if that noise floor is modulated at very high SPL. PWK once commented on the unnaturally "ghost-like quiet" in digital recordings of the 80s. By way of analog comparison, flat screen TVs and digital camera contrast ratios cannot begin to approach the ~13.5 f-stops of human vision dynamic range. They're more like 8.5 f-stops.

I believe 90 dB noise floor (linear scale) is good enough. The DSP noise in the Crown XTi-1000 was really objectionable when quiescent, but that noise perception disappeared when anything was playing. If it is only the noise floor that increases when the dynamic range rarely exceeds 90 dB, then I'm comfortable with that, and I believe that I can work around it in gain programming. However, if someone wants to hand me a Yamaha digital crossover at a fall-off-the-truck price, I'd probably consider it seriously.

I find that it's the other distortion types/sources that are objectionable - such as that in passive crossover networks undergoing Ohmic heating, cooking inside of closed or nearly-closed boxes, and other forms of electro-magnetic distortion. In fact, I'd probably start to examine/invest in solutions for magnetic distortion dynamics of acoustic drivers, time misalignment of drivers in all my loudspeakers, and diffraction of dual-mouth corner horns before I'd get excited about dealing with 90 dB electronics noise floor.

For me, it's all about system-level trades.

My $0.02.

Chris [:#]

[Guest " "]

Going to buy 4 more la scalas tomorrow. That will make 9 total. Nine point one surround anyone? It's hard to pass up good deals on these speakers.

This is good.

[Coytee]

That will make 9 total

Is the beginning of the LaScala Cartel?

[ClaudeJ1]

Putting it back in perspective, the distortion of your speakers will dominate the distortion from the Dx38, but it's just one of those things that is accumulative and ultimately not necessary with more modern units.

So what is the list of the more modern units? Does the EV DC One fit that bill? How about the Ashleys? or, gulp, the lowly Behringer DCS 2496?

[ClaudeJ1]

(Don't get me wrong -- it is still a very, very good sounding converter. It's just that more modern designs sound even better.)

Greg

OK, so which units do you consider a "modern design."

[ClaudeJ1]

Anyone using AES/EBU digital inputs in their digital Xovers and bypassing the A/D conversion??

[Chris A]

Going to buy 4 more la scalas tomorrow. That will make 9 total. Nine point one surround anyone? It's hard to pass up good deals on these speakers.

Wow - you must have a really big room...[] Enjoy!

Chris

[Chris A]

Anyone using AES/EBU digital inputs in their digital Xovers and bypassing the A/D conversion??

I asked a related question last year on another forum about that exact subject. I didn't get a very good answer, IMHO.

Chris

P.S. see the following -- http://www.sweetwater.com/store/detail/ODL312/

[Edgar]

OK, so which units do you consider a "modern design."

Anything that does not use a dual-range ADC. Modern monolithic converters are achieving very high dynamic range and SNR, and good sound, without the need for range switching.