Ayre DX-5 bluray player - Page 29

It probably could be done, but space is *extremely* tight inside the DX-5. You might have to leave the top cover off or something! But the question is whether it would be worth it. The main Oppo processor has an ARM core and it provides the conversion of DSD to PCM at 88.2/24. I don't know how much improvement there would be in going to 176.4/24 and I don't know whether the Oppo processor has better sounding filters than the DVD Upgrades board or the other way around.

If you pop the top off the unit and look at the board, you will probably find a chip made by NPC in there. They have made two or three models of DSD-to-PCM conversion chips. However DSD has a lot of high frequency noise just outside the audio band. If this is not filtered out, it can cause a lot of problems for the downstream components. Sony sets a minimum standard for the amount of filtering that must be applied. In many cases, there is virtually no difference in the results between converting to 88.2 kHz and 176.4 kHz.

Bottom line -- leave the board alone, be happy with the conversion in the Oppo processor.

Charles,
Thanks very much for your thorough response. That helps me think about the decision. Just to clarify, my statement about comparing the USB DAC to the CD player referred to my reading of the Stereophile review, not my own judgment. In my system, the QB-9 playing even 16/44 files sounds "significantly" better than the CX-7eMP and the CX-7eMP is no slouch. I was impressed by the sound quality of both CD and SACD on the DX-5 I heard in my local shop (Music Lovers Audio-SF)-"significantly", perhaps even "dramatically" better than the CX-7eMP-but haven't yet been able to convince myself about the DAC. I look forward to reading the reviews and to further auditioning for myself.
Thanks again,
Rick

Probably spectaculaly good advice - and given my technical limitations, unquestionably the way that I will go. It is fun to speculate on this stuff however. Thanks for your considered responses.

Dug this up ca1ore ;excuse the need to translate the german ;

https://www.dvdupgrades.ch/product/S...ard/24308.html

Interesting discussion here; http://www.head-fi.org/forum/thread/...-pcm-converter

Charles could best discern the pcb layouts effectiveness [or maybe its ic quality ?]

Ah, yes, you found it. That is the board that I have installed in a Denon player. I will have to dust it off and do a 1:1 comaprison with DSD/PCM in the Oppo.

That mirrors my experience exactly. I love the CX-7eMP and think it is a great CD player. But the QB-9 has more dynamic "life" and is just more fun to listen to. But I also think that the whole computer audio vs. CD thing is a huge paradigm shift and I completely understand it when people don't want to get involved with the whole process of trying to connect a computer to their music system. I think it makes a lot more sense when you use a computer every day for work and are very familiar with them. Otherwise just stick to a normal disc player.


The fact that you enjoyed SACD on the DX-5 just goes to show that reviews are pretty limited. The most they can do is to point you towards something so that you can check it out for yourself.

If you have a QB-9, it should be a simple matter to borrow the DX-5 from Music Lover's and substitute it into your existing system. I think you will hear a noticeable, but not significant improvement, but there's only one way to find out!

I think the real deciding factor is if you want to use the DX-5 for video. If so, then it is a no brainer as it will do everything better than what you have now and if you stick a flat screen (preferably a plasma) on the wall between your speakers, you instantly have a 2.0 channel home theater with incredible sound.

It's an odd duck. The first thing is that all of the heavy work is done by the large chip on the bottom. That is an FPGA, which is programmable to do almost anything. They chose an Actel. Traditionally Actel used a technology called "anti-fuse" where programming the device literally involved blowing small fuses inside that connected things together. The programming machine would blow all the fuses that you didn't need for connections.

The disadvantage is that once you program it, you can never reprogram it. This is especially painful when you are developing the program and you might go through dozens of iterations until everything is working properly. And this is not a cheap part -- the only price I could find in a quick search is $43!

However, this is a new generation of Actels. It no longer uses anti-fuse technology, but instead uses FET switches to connect the elements together. The FET switches work like SRAM so that the device needs to be reprogrammed each time it is turned on.

Other companies that make FPGAs put the program in a separate chip that connects to the FPGA. But DVD Upgrades is very secretive and protective of their designs. (I think they were ripped off a lot during the early days of "codefree DVD" and became very paranoid.) So this Actel puts the programming chip inside the main chip. That way the program cannot be intercepted and copied. The programming chip inside uses flash memory and can be re-programmed with a special machine.

On the top side of the PCB is an AKM chip that is apparently used as the S/PDIF transmitters. It is strange that they didn't just use the FPGA to do this, as it is a trivial function. I don't understand that at all. Then there is a voltage regulator and one more chip I can't identify. It is probably a MUX to select either the PCM input or the DSD input.

At any rate, this all means that they are designing their own filters for DSD-to-PCM conversion. I don't think the Actel parts include much support for DSP work, so I would think the filters would have to be quite simple, especially as all six must fit into the one part.

But the proof is in the pudding. A listening test would be very interesting. The problem is that there is no way to do a fair one. The DX-5 can only output the data from an SACD on the HDMI connector while the DVD Upgrades board uses S/PDIF. So a lot would depend on how well those two inputs are implemented on the surround processor used. But it would be interesting for a rainy Saturday afternoon. Let us know the results if you try it, ca1ore.

I'll do it over the holidays and report back. No way to truly isolate the effect of the board, however, as it will be 'Denon/DVDUpgrades versus Oppo'.

Simon

That's not so bad, but it will also be "HDMI vs S/PDIF", which will probably make more difference. But it will still be interesting no matter what. And the other thing is you can change between 88.2 and 176.4 kHz (don't forget to cycle the power between changes) to see how much sonic difference there is.

The NPC chip that converts DSD to PCM has almost identical frequency responses for both sampling rates. In both cases they have to roll off the response above 50 kHz to keep the out-of-band noise that is on the disc from reaching the rest of the system. I have no idea what the DVD Upgrades algorithm is like.

Edit: I read one of the linked reviews for the DVD Upgrades device. It was in Italian, so I used Google translate. As near as I could tell, when set to 88.2 kHz, that device rolls off at 30 kHz and changing to 176.4 extends it to 50 kHz. I haven't measured it, but would assume that the Oppo processing engine has a sharp filter around 42 kHz or so. That would be standard digital practice.

Yes, except I don't think my Meridian 861v6 will accept 176.4. I confess I'm not sure, will have to check. I'm also using a Meridian HD621 which does seem to reduce SPDIF vs. HDMI differences, but you are probably right about that making more of a difference. I'll do it report back.

The 861 will not accept anything higher than 24/96. The HD621 can be set to output 24/96 to the 861 and it will down-sampled anything higher to 24/96 as well as upsample anything lower.

So, I decided that this whole thing is just too much fun to wait to the holidays, so I played around a bit last night ....

My comparison streams are the Denon DVD-3910 with DVDUpgrades DSD-to-PCM conversion board, outputting via 3xSPDIF to my Meridian 861v6, versus, Oppo BDP-83 (with internal DSD-to-PCM conversion) via HDMI to a Meridian HD621 and then digital via MHR on RJ45 to the 861v6. Not a very effective way to isolate the performance of the DVDUpgrades board, but it is the best I can do.

I need to spend a lot more tme listening, but my initial impression is that the Oppo is superior. Bass was clearer and more extended, and the sound seemed more dynamic. Curiously though, from the Denon my 861v6 displayed 'Mov6 88k', whearas from the Oppo it displayed 'Mus6 88K'. Going to have to break out the manual (sigh) to try to figure out what this means and if this is biasing my results.

Assuming my views persist, it will be interesting to see if the DX-5, over the HDMI Audio output is, in turn, superior to the Oppo (I have the Jan Sterophile, but have not yet read Kal's impressions).

Simon

Considering hdmi is supposed to be worse than sp dif with jitter Simon ; speaks well of the oppo let alone the dx5
It looks like a dsp mode is in the mix ; this is from a G68 ;

It is, especially if you are using any bass management as the 861 has different setups for Mus6 and Mov6.

BTW, I did a similar comparison between the Oppo via 3xS/PDIF vs. via HDMI/HD621 when I got the HD621. http://www.stereophile.com/content/m...ound-38-page-3

Kal

Yes, I'm not sure why 861 is defaulting to 'Mov6 88K' for the Denon, but 'Mus6 88K' for the Oppo - and whether it makes any difference? I have a well-developed case of 'anti-manual-itis', but I guess I will have to go and take a look.

I'm talking about what 861 is displaying as its input, not what DSP I am using for the output. In the latter case, I am using the same 'Discrete' output DSP with common bass management. Does 'Mus6' vs 'Mov6' on the input make a difference? I ought to know I suppose

Movie and Music have different crossover slopes. Not sure if that is in the manual. There was some active discussion of this at Hitchhiker's guide and how to fix the defaults.

I ought to know the answer to this question, but does the 621 accept S/PDIF signals?

If so, then you would be able to create a much fairer comparison. The problem is that the 621 has it's own set of digital filters built in. So the result is a concatenation of the external filters in the 621 and the internal filters in the 861.

Concatenating digital filters is nothing unusual. In fact it is the rare product that doesn't have a concatenation of filters. But when concatenating filters, the first filter has virtually all of the sonic impact. So by adding the 621 you have changed the sound of the system.

The 621 is kind of an odd thing. It "upsamples" (a silly term, used only for marketing and not by engineers) 44 and 48 kHz signals to 88 and 96 kHz. It "downsamples" 176 and 192 kHz signals to 88 and 96 kHz. The digital filters are supposed to be the minimum phase "apodizing" filters, so I suppose that it performs the same "apodizing" function on incoming 88 and 96 kHz signals.

It's a different approach than I would take, and perhaps a different approach than Meridian would take if they were starting from scratch. The nice thing about the 621 is that it allows you to add HDMI capability and also presumably improved digital filters (but probably only on the HDMI input) without having to buy a complete new processor.

Let us know what you find out. This is an interesting experiment!

The 621 does not accept S/PDIF signals - just HDMI. However, my Oppo player does have a 4xS/PDIF ouput board that was added as an after-market mod (used prior to my getting the 621) and I do not believe it has any kind of digital filtering, so I will try that also which will enable me to bypass the 621.

Simon

Actually the apodising (because Meridian is English) filters only work on 44.1 and 48 kHz signals. Can one actually apply an apodizing filter and not upsample?

Yes.

We tried all kinds of things when we were experimenting with digital filters in preparation for the QB-9. There were some things we liked about non-oversampling and some things we liked about apodizing filters. So we decided to combine the two.

Basically what you create is a notch filter, with the sample rate unchanged. The notch is at Fs/2, which is where all the ringing is created by the half-band filters that are upstream in the recording chain. The notch should start at 19 or 20 kHz and then reach a maximum at 22.05 kHz (for CD). Then because of the way that digital filters work, it will rise back up to the reference level at 24 or 25 kHz.

The basic pattern will be repeated into infinity in the mathematical world inside the digital filter. However in the real world the "hold" function of the DAC chip will superpose a sinc function in the response. In addition there will be some analog filtering on top of that.

But it turned out to be a clever idea that didn't pay off. It sounded less good that other approaches. We put the best sounding filter into our "Listen" switch, and the best sounding filter that had flat frequency response in the "Measure" position.

That will be the fairest way to compare the DSD-to-PCM conversion of the Oppo versus the DVD Upgrades board. I think that the digital filter in the 621 will affect the sound to a noticeable degree. Thanks for your hard work on this. I am very curious as to the outcome.

Some of the new Sony Blu-ray players will play SACDs. You can send the data over the HDMI output either as DSD or 176.4 kHz PCM. But I don't think 176.4 PCM will be much different than 88.2 kHz PCM for this application. In both cases you need to start rolling off the frequency response above 50 kHz in order to keep all of the out-of-band noise that is inherent to DSD out of the rest of your system.

In my case the problem was due to the delay of the picture caused by the AV-pre compared to the analog XLR-outputs of the DX-5. Switching off the video conversion circuit within the AV Pre solved the problem.

For those who may have tried it, does the Ayre share the BDP-83 clipping problem on the 7.1 soundtrack from Toy Story 3?

Hey, that's great! I'm glad you solved the problem. Let us know how everything else goes.

I also have an idea to solve the problem that the HDMI Audio output does not force the DSD signal into PCM as necessary for the Marantz AV7005. It's very simple: I recognized that once the DX-5 was forced to PCM by switching on the AV7005 and set it to the appropiate input, the DX-5 remains in PCM unless the tray is opened or the the SACD stereo and multichannel layers are toggled by the audio button on the remote control. Thus I will simply connect the DX-5 using two HDMI cables into separate HDMI inputs of the AV7005, "activate" PCM by switching the Marantz to the input connected to the HDMI A/V output of the Ayre and then switch the AV7005 to the HDMI Audio output of the DX-5 and PCM will remain. The HDMI A/V output connection is needed anyway for the picture, so its only an additional HDMI cable...

Merry christmas and a good 2011 to you and everybody reading this!

I guess I'm a little confused. Why not just set the DX-5 for PCM output using the On-Screen Setup Menu?

Is there a situation when you want to listen to DSD on a different preamp? Maybe the analog outputs?

~~~~~~~~~~

On a different topic, the ideal situation would be to send the video to the display and the audio to the surround processor. This allows for total electrical isolation between the display and the audio system, which improves both the picture and sound quality.

The problem is that most people have a second source (satellite or cable). If you don't care about the sound quality and your display has speakers, you can send the second source there. But it's not that practical for most people. Someday we will need to make a HDMI switcher that has an isolated audio output. Then you can connect all your sources to it and it would isolate the video and audio systems. I haven't gotten around to it because all I watch is DVD or Blu-ray. I don't have cable or satellite and never will.

First, I don't want to switch on the TV set or the beamer when listening to music only.

Second, normally I listen to two channel music by the DX-5's XLR output feeding a Mark Levinson No. 326S and a pair of No. 436. In case of SACD two channel, I prefer DSD. Only for multichannel purposes (SACDs and BluRays/DVDs) the separate AV preamp (Marantz AV7005) comes into the game. It is also connected to the ML326S to reproduce the main channels in the multichannel layout. Center and surround signals are amplified by a separate Marantz MM7055 5 channel power amp. The AV 7005 does not accept DSD signals, thus PCM conversion is required for multichannel SACD listening.

I use overall 5 video sources, including the DX-5. They are all HDMI-connected to the AV7005, which serves as a switch, a scaler, and a multichannel decoder and in addition distributes the signal either to my beamer or to my TV set. Connecting the DX-5's HDMI Audio output to the AV7005 and its HDMI A/V output to the beamer directly would result in best audio and video quality but skip the option to watch video signals from the DX-5 on the TV, unless I install an additional HDMI switch, which I would like to avoid. Maybe I use the composite video output of the DX-5 for connection to the AV7005, just for purposes of usage of the on-screen menus of the DX-5. I will think on it...

Yes, it is somewhat annoying to have to do so. With the C-5xeMP, one can navigate most any disc without the need for a display. But since the DX-5 will usually be connected to a display, there are some discs that require the display for proper navigation. However, once you have found the proper starting point, you should be able to turn the display off while listening to the rest of the disc.


That is the best way to connect it for multi-channel music.

The one thing that will tend to degrade the sound of your system is the fact that the Marantz is connected to your display. If you are not using the display while listening to music, you can improve the sound of the system by plugging the display into a switched outlet strip. Normally the display has a switching power supply that is running even when the display is in "Standby" mode. By turning the switched outlet strip off, the switching power supply will also turn off. However, you will need to turn it on manually.

Also, if you have a connection to a cable system, I would recommend using an isolation transformer such as the Jensen Model VRD-1FF:

http://www.jensen-transformers.com/iso_vid.html


I don't know what the other 4 sources are, but I am confident that the DX-5 will provide the best picture quality. However, the Marantz has the Anchor Bay ABT-2015 scaling chip. This is essentially the same chip as is used in the DX-5 but is on a newer (90 nm?) process that requires no heatsink or cooling fan. So it will do a good job with many sources. Inside the DX-5 it has the advantage of having the frame cues transmitted to it directly rather than having to try and interpolate to figure out which frames are which.

But as soon as you connect another video source to both the display and the audio system, then the audio isolation is shorted out. It's not ideal, but you will still be better off using the DX-5. Both the video quality and audio quality will be better than your other sources. I would definitely not recommend relegating the video output of the DX-5 to a composite monitor!

Most displays have multiple HDMI inputs these days. That would solve one problem. I don't know if you would gain anything by using an external HDMI switch over the one that is built into the Marantz. They are inexpensive enough that it might be worth a try just to see if it looks or sounds any better.

Please keep us all posted as to the results of your experiments!

I do this to navigate something like the DVD-A menu or changing the set-up. I use a small 7" alpine display intended for a car and it just sits on a shelf with my CDs. This keeps me from turning on my PJ unnecessarily.

Could the composite also negatively effect the audio quality - I haven't experimented with it...or the HDMI for that matter. I assume you made the above comment because the HD PQ is so good.