HDX as UPnP server - effect of external power supply?
Posted by: rjstaines on 03 March 2012
My HDX is the preferred UPnP server to my NDX. Would the HDX's UPnP performance be improved by adding a power supply, or is it only the analogue side of the HDX that benefits from external power?
Reason for asking is this: currently have NDX / XP5XS into DAC/555PS (into 252/300), fed by HDX UPnP. I have ordered an NDS and plan to sell NDX and DAC, keep 555PS of course for NDS, but then the XP5XS - sell or add to HDX ??
One is not flush with cash, so selling the XP5XS is favourite, but for me SQ usually overrides financial considerations (probably why I'm always broke!).
I would like to call on the combined wisdom of the forum on this matter, thanks gentlemen (and ladies?).
Roger
This also answers the question if trading in a HDX for a US as as UPnP would have negative SQ consequenses; clearly no. The US is rather more still an overkill.
Could have a point... NDS with 555PS AND an HDX too... may be a bit overly redundant and ridiculous. Perhaps I'll dig out my one-box Dansette, it holds up to 6 LPs you know. There again, on reflection, I find nothing wrong in being overly redundant and ridiculous, ...so long as it sounds nice.
No trouble at all - on the contrary. All rips are the same, if they are bit perfect. And that is possible even with the cheapest notebook.
Possible, but IMHO not probable.
Gentlemen, this has nothing to do with MHO. You may well belive that a simple notebook is not capable of producing a bit perfect rip, but really this is not a matter of believe, it is a fact.
You may further believe that a bit perfect rip produced by a HDX will sound different to one produced on besaid cheap notebook. This is then the best positive feedback that the marketing department of Naim can wish for, and it is the typical reaction of a digital immigrant.
Roger,
sorry for having been quite rude on thIs. It is of course your own decision what you like or do not like. I am just trying to bring some rational into this.
After all, this forum is all about believe and misbelieve, isn't it?
Lets continue .....
Should a streamer land on our Fraim our HDX-SSD won't be going anywhere. Yes it is an expensive option, but you could argue that the NDS is incredibly expensive for essentially a digital out socket. Each to our own. It's a great ripper, idiot proof to set, use and run, and it can even play CDs. How charmingly retro.
Like I said: Digital imigrants on the move ..
Well, if one doesn't feel comforable with PCs, the HDX may be a viable solution. This is what it was designed for, ripping and serving for poeple without computer affinity.
I would recommend stepping away from the geekery and spending of money and actualy listen to some music.
Jesus. Ethernet cable is ethernet cable. If it delivers the signal then it works. Power supply quality for a upnp server, s******.
Step back, take a look at what you are saying, and consider how the rest of the world would view us.
I looked at this 'difference between rips myth' last year with a few on this forum, as vodoo magic was entering into the discussion and confusing people un neccessarily...
We decompiled right down to byte level several rips using specialist decompiler tools. We looked at Naim, EAC, dBpoweramp and itunes. We asked each ripping engine to create a WAV file, where we had a choice.
Naim and dBpoweramp produced extensible WAV files. iTunes and EAC produced the older legacy canonical WAV file format. (developed in the 1980s). All wave files have a DATA chunk. This is is where the raw rip or sample data is stored. This was IDENTICAL on all rips. Ie across a file of tens of millions of PCM samples there was no difference WHAT SO EVER to any of the values. The sample data was also contiguous, and so not chopped up, and it's length was IDENTICAL.
Differences in the WAV file were down to header type (based on WAV type) and whether it contained INFO list and/ or ID3 chunks. But these ate completely seperate and disconnected to the sample data.
We also challenged forum members to discern any difference between WAV files (as a blind test). Unsurprisingly no one could tell any difference between them.
Therefore we concluded all valid rips are THE SAME.
The differences in the ripping engines however affect usability but NOT sound quality. For example
1) interface (computer or not)
2) ease of use collecting meta data and album art
3) ease of editing the above
4) strategy for handling damaged CDs.
5) ability to store metadata in WAV files
6) file types supported by ripper.
7) ability to cross check checksum of rip, ( to determine faulty CD)
Simon
+1 x 100 to Simon's excellent post.
I have tried less extensive experiments than Simon, but we could add CD Paranoia to the list.
This also makes perfect rips.
There is nothing wrong with using a UServe or HDX as your UPnP server/ripper though. Both are well made Naim components that do what they say they will. Naim rippers cannot rip better than perfect, but that is not a criticism - just a fact.
With HDCDs, dBPowerAmp remains the best option (AFAIK) because it decodes the extra data and puts it in the rip, some other rippers simply copy the HDCD code and rely on your DAC decoding it, which my Naim DAC cannot.
Another interesting test is to record the analogue out from two rips or two different CODECs - invert one and then play back the two simultaneously then all you'll hear is the difference. One studio that tried this reported no differences. However, you would have to accept that the way the experiment was conducted was sound enough to support the claim. I also don't think you can prove anything this way because you cannot be sure that different music would not give different results.
However I concur with the studio experiments because to me all error free rips sound the same and all lossless CODECs sound the same using the same song on the systems I have. I would be disappointed if the NDS could not extract the PCM for any lossless format it supported with equal aplomb.
So I would say using the HDX as the UPnP will work very well and give superb result, using at a CD ripper will work and give superb results. I cannot see any way that adding a power supply to the HDX would change this, but I would suggest trying isolating the power to HDX from the power to the rest of your system, as this may have an effect. Be interested to know if anybody has tried this?
Guido,
what do you mean by "extra data" that dbpa is putting into the rip?
@PinkHamster - if dBPoweramp detects that the CD is HDCD enoded (whichi is a special DSP that simulates extra sample word length) then it uses (or so I understand) a freely available 'HDCD -like' utility that simulates the HDCD DSP algorithm and expands the sample length (from my memory) from 16 bit to 20bit. DBPoweramp then truncates the last 4 bits, and encodes as 24 bit file. By default DBPoweramp does NOT do any of this unless you enable it in the DSP section of the tool. If not enabled any HDCD encoded PCM pass un adjusted into the WAV file, just as if the CD player or ripper did not revognise HDCD.
I have a few HDCDs and they do rip rather well into 24 bit files, and sound rather good, albeit with HDCD the noise floor is raised.
Simon
I see. Thank you, Simon.
As I understand it
A Doom Nine programmer wrote hdcd.exe to extract and decode HDCD data in 16-bit WAV file ripped from a HDCD disc. You get 24-bit WAV output files with four bits of padding per sample. They did not put this in the public domain as they wanted to make money out of it. One of dBpowerAmp's DSP effects is a front-end to hdcd.exe. You get 20-bits of PCM data in the file rather than the usual 16.
I've not checked this, but HDCD ripped this way do seem to sound crisper and more detailed. I have 5 such CDs that a friend ripped for me, as it required Windows which I don't have. I thought I had dBPA running on my Mac, but it won't rip CDs, it will only do file conversions under a Wine emulator.
@Roger & scillyisles
Thank you for having had a look. Apparently nobody tried. Funny because it is quite a subject when using SqueezeBoxes. Logitech recommends the "native" setting too, but not really on a sound quality basis, but more because the Logitech Media Server adapts the transcoding to the capacity of the client. Others tried to transcode to LPCM on the server when using a SB Touch for instance, and found some improvement.
Impressive gears of yours TC ! Are you satisfied with the replication of your QNAP NAS ? Did you try the encrypted replication as well by any chance ? (for more sensitive data obviously)
Maurice
The NAS replication between the two music libraries works extremely well. I brought the 2 NAS units together for the initial replication as it was going to be a ton of bits...about 760 albums. The process is now automatic...whenever I rip a new CD or download an HD file...the NAS that gets the new data automatically kicks off the replication process. Configuring the DynDNS and router port mapping was very easy. I don't have any encryption set up. Each NAS has a 4 Barracuda XT 2TB drives in raid 5 for my music and one 2TB drive as a separate share for Time Machine backups of my Mac computers. I use QNAP 559 pro 2 NAS models...love them
Thank you TC.
ATB
Maurice