It surprises me that there is no other product that does digi in/out only (i.e. without the analogue) crossover. With the abundance of cheap and god D/A it would seem like a natutral product.
Over sp/dif, it is the source that sets the level of jitter as lot of DACs only have basic clock recovery, no real jitter supression in the audio range. A good DAC of cource does this - e.g. Benchmark DAC-1.
oettle said:
Different applications cause different needs. I'm using DCX in stereo mode for two 3-way speakers and my goal is max sonical quality. Using DCX as a processor for multi channel applications digital outputs might be a need and if well done better than a D/A and an A/D conversion.
I've to stop now because I've to bring the Chritmas tree in place for my children.
Merry Christmas (Happy holidays) to all
I've to stop now because I've to bring the Chritmas tree in place for my children.
Merry Christmas (Happy holidays) to all
Hi,
I'm finally getting around to installing Franks digital in / resampler mod. As I mentioned before the goal is to not only install the mod but also try to make objective evaluation and measurements.
I'm doing the preparations today so if anyone has additional ideas what to include in this test then do tell 😉
Currently the plan is to (pre and post mod):
*Make a thd measurement with 44.1k and 48k sampling
*Make the ARTA multitone measurement with 44.1k input signal
*Make the Jitter test with Dunn signal with 44.1k input sampling
*record a piece of music pre and after mod and use Liberty Instruments Diffmaker software to analyse for differences.
Regards,
Ergo
I'm finally getting around to installing Franks digital in / resampler mod. As I mentioned before the goal is to not only install the mod but also try to make objective evaluation and measurements.
I'm doing the preparations today so if anyone has additional ideas what to include in this test then do tell 😉
Currently the plan is to (pre and post mod):
*Make a thd measurement with 44.1k and 48k sampling
*Make the ARTA multitone measurement with 44.1k input signal
*Make the Jitter test with Dunn signal with 44.1k input sampling
*record a piece of music pre and after mod and use Liberty Instruments Diffmaker software to analyse for differences.
Regards,
Ergo
oettle said:
Frank,
There is no requirement to have a SRC on the receiving end. It is possible to simply accept the audio at the rate sent over the S/PDIF connection and work with it from there. Feeding the XR55 over S/PDIF from the Behringer stays at 96/24 for example.
Even if one had a SRC on the receiving end that does not make S/PDIF a lossy transport. The transport transported the data, by the time the SRC gets the data you are already back to I2S. S/PDIF simply transported it, the SRC is what is resampling the data, not the transport.
Shawn
Hi, The Christmas stuff is done.
Shawn,
Where does your DAC get its clock from? I assume you are using a PLL on the receiver side which generates e.g. a 12 MHz master clock from a 96 kHz data stream. That's a max jitter solution! I even would prefer an additional SRC and reclocking the data.
If you are aware of an ASRC which isn't lossy please tell me the part number. Looking in the data sheets of the ASRCs I know they are all talking about resolution, distortion, .... A lossless ASRC wouldn't need any dB value in its data sheet.
TNT,
This Benchmark DAC isn't only a DAC it's a PLL, a SRC, a clock and a DAC all together. It's down sampling data from 96 kHz to 52 kHz and it adds cost and jitter.
Frank
Shawn,
Where does your DAC get its clock from? I assume you are using a PLL on the receiver side which generates e.g. a 12 MHz master clock from a 96 kHz data stream. That's a max jitter solution! I even would prefer an additional SRC and reclocking the data.
If you are aware of an ASRC which isn't lossy please tell me the part number. Looking in the data sheets of the ASRCs I know they are all talking about resolution, distortion, .... A lossless ASRC wouldn't need any dB value in its data sheet.
TNT,
This Benchmark DAC isn't only a DAC it's a PLL, a SRC, a clock and a DAC all together. It's down sampling data from 96 kHz to 52 kHz and it adds cost and jitter.
Frank
Frank,
"Where does your DAC get its clock from? "
When I am done with the next step I will have no DAC after the Behringer, or at least not a PCM DAC. The receiving end can either use that sources clocks in slave mode or be set to recover the clocks from the S/PDIF input. The receiver chips I'm using is a very low jitter part on a recovered clock. From there the data basically gets buffered/reclocked (not resampled) anyway through the action of the DSP it passes through.
It gets rid of an unneeded D/A, analog output stage, analog input stage and A/Ds.
"If you are aware of an ASRC which isn't lossy please tell me the part number. Looking in the data sheets of the ASRCs I know they are all talking about resolution, distortion, .... A lossless ASRC wouldn't need any dB value in its data sheet."
I never said a SRC wasn't lossless. I said a SRC isn't part of S/PDIF transmission. Two different processes. DACs aren't lossless either since they have the same measurements in their data sheets and of course most DACs basically have SRC internally as well. The AKMs in the Behringer certainly do for example.
"This Benchmark DAC isn't only a DAC it's a PLL, a SRC, a clock and a DAC all together. It's down sampling data from 96 kHz to 52 kHz and it adds cost and jitter."
The Benchmark is very resistant to jitter on the S/PDIF transport. Real world data not theory.
Shawn
"Where does your DAC get its clock from? "
When I am done with the next step I will have no DAC after the Behringer, or at least not a PCM DAC. The receiving end can either use that sources clocks in slave mode or be set to recover the clocks from the S/PDIF input. The receiver chips I'm using is a very low jitter part on a recovered clock. From there the data basically gets buffered/reclocked (not resampled) anyway through the action of the DSP it passes through.
It gets rid of an unneeded D/A, analog output stage, analog input stage and A/Ds.
"If you are aware of an ASRC which isn't lossy please tell me the part number. Looking in the data sheets of the ASRCs I know they are all talking about resolution, distortion, .... A lossless ASRC wouldn't need any dB value in its data sheet."
I never said a SRC wasn't lossless. I said a SRC isn't part of S/PDIF transmission. Two different processes. DACs aren't lossless either since they have the same measurements in their data sheets and of course most DACs basically have SRC internally as well. The AKMs in the Behringer certainly do for example.
"This Benchmark DAC isn't only a DAC it's a PLL, a SRC, a clock and a DAC all together. It's down sampling data from 96 kHz to 52 kHz and it adds cost and jitter."
The Benchmark is very resistant to jitter on the S/PDIF transport. Real world data not theory.
Shawn
Haaa, What shall I say?
Unfortunately real world is so complex that it needs so many theories that nobody is able to know them all (including myself). But our brains are built for learning a whole life and evolution gave us this wonderful incentive system of good feelings when learning. For most of us it's a problem to admit errors. But there is no learning at all without making errors. Evolution is the best example for this. So let us having fun by learning from each other here on this forum.
Frank
Unfortunately real world is so complex that it needs so many theories that nobody is able to know them all (including myself). But our brains are built for learning a whole life and evolution gave us this wonderful incentive system of good feelings when learning. For most of us it's a problem to admit errors. But there is no learning at all without making errors. Evolution is the best example for this. So let us having fun by learning from each other here on this forum.
Frank
I may be missing something here, but if you wanted a say 5 way speaker, wouldn't it make more sense to split the digital signal before it gets to the DCX's? I can understand the digital out if you needed to get more EQ for example (but I would think that would be better corrected by other methods).
If this first DCX does all I want, then I'm thinking of moving up to a 5 way.
If this first DCX does all I want, then I'm thinking of moving up to a 5 way.
Start with the basics: better drivers and design. If you need a lot of EQ in any speaker, irrespective of the xover design, then you're starting with the wrong parts.AR2 said:
Brett said:
Yes, but you may be able to do it with DSP eq for a fraction of the cost a 'flat' speaker would set you back. Also, with DSP eq you can correct for room effects, which is pretty much impossible otherwise.
Jan Didden
Room EQ, except at LF is a bit of a waste in the implementations I've heard so far. They've all seemed too locked in to one position, and I don't do the head in a vice thing well.janneman said:
I can't think of a driver I've used in ages that needed extensive EQ.
Besides adding extra O/P boards and other componentry can bring the cost up to the point where the better basic driver worked in the first place. My engineering always taught to get it right from first principles and only correct as neccessary. Never let me down yet.
Now it seems to be me who should learn something. Brett, can you explain a German guy with only little English skills what's the basic massage of your posts?
So far I understood you prefer better drivers than using a lot of EQs and you do not like room correction. But how do you realize crossover and what is the benefit of a 5-way speaker?
Frank
So far I understood you prefer better drivers than using a lot of EQs and you do not like room correction. But how do you realize crossover and what is the benefit of a 5-way speaker?
Frank
That's about it. I'm not convinced about room correction as it seems to me that it corrects for one point in space. So if you move your head, some of the benefit goes.oettle said:
Regarding drivers and general speaker design, I have a strong preference for drivers that behave well out of band, and those that don't need much if any EQ. Start with the best materials, rather than use something lesser and try to "fix" it by some other technique.
IME, I don't like to use a driver over more than a decade (10x) frequency range. This is true for horns and direct radiators. I also prefer 20-60Hz to be handled by dedicated subs. That leaves about 8 octaves to be divided up between the remaining 4 drivers. These drivers can then be selected and used so that they have the desired frquency response, low distortion, directivity etc that I choose. They will also be well behaved outside of the desired frequemcy band so with good filtering (say LR24 via DCX) any problems well outside band will be well attenuated.oettle said:
I like horns, so when I have the chance, I'd like to try a system such as the one below> Please note, this is only a thought exercise at the moment.
HF: B&C DE10 on a DDS 90* flare >2kHz
MF: JBL 2485 on conical 700-2k
LM: 12NDA520 on trax 200-700
MB; 2x15 hypex (already rested design) 60-200
LF: 4 x 3015LF tapped horns
Two DCX would allow for each TH to be spaced (and delayed) correctly in the room to minimise nodes. They would also allow path length differences to be compensated for. There should only be minor EQ in each drivers passband and each can operate a bit above and below where I'd set the xover points.
This design is for a very large open plan room that has held about 150 people at a party. Low distortion, no dynamic compression are two of my primary goals.
In the short term, I will be 3 way with the DCX > 60Hz fed digitally, and an opamp based LPF for <60Hz to the subs fed from the analog out of the player.
I do not like subs, and yet I am to see bass driver to cover 20 - 200 without serious EQing. Just a few subs in my mind sound good, but for the sound I am interested in - two channel music only, they are not needed. The only way to avoid Eq is multiplying bass drivers, but than lets say with 8 12" or 15 " (4 per side) we are talking about esthetics that is limited to few environments.
I don't like speakers that are EQ'd extensively to get a flat response adding a ton of distortion well into the midrange.AR2 said:
It depends on what you listen to as to the amount of LF present.
Brett said:
Well, YMMV of course, but I have yet to see a room that doesn't mess up any good speaker by the time the vibrations reach your ears. (Come to think of it, I do know one room that doesn't. Saw it in Indonesia, cost about half a million bucks).
I average room equalization over several positions near the listening position, and for me it works. And your 'head in a vice' remark is a strawman really; whether you use RE or the most expensive speakers you can buy, it's no different. The sound field *always* changes when you move your head.
Jan Didden
Brett said:
Brett where does that distortion come from? Are you talking about the electronics that do the eq?
Jan Didden