Rip your files the NAIM way...or else???
Posted by: Geoff P on 01 August 2012
On another thread here which discussed limitations to file transfer and storage on the UnitiServe outside of ripping in the box (and by reference also the HDX/NS0x series swell) the following responses were given to questions suggesting bit perfect rips by commercial software such as dB Poweramp or EAC should be accessible and transferrable with the same results as the internal Naim controlled rips.
Paul Stephenson:-
"Easy naim rip via our server easy db and eac good but our rip and our drive choice usually outperforms"
Phil Harris:-
"The point here is that the UnitiServe (and HDX / NS0x) are designed as a hardware and software combination that gives a good set of results without the end user having to work out and put together their own combinations of hardware - i.e. the rips should be as good as it is possible for them to be - bit perfect.
EAC or db *CAN* also achieve bit perfect rips but EAC or db are only part of the solution - there's hardware that they sit on top of and that too has to support the process. Just throwing EAC or db at a CD on its own does not guarantee a bit perfect rip if the hardware tey're running on doesn't support that.
If there are differences in the rip then there *WILL* be differences in the payback of those rips and if a rip produced by EAC or db isn't bit perfect then there's nothing that any utility written by anyone can do to make them so...
Paul Again:-
"....as Phil replies its the combination of hardware and software, the files will look identical from,db,eac or itunes but from the experiences we have had here the sound is not."
The issues for the larger digital ripping community that this raises should be answered if the above statements are taken to be correct.
1) By implication the US has a hard disk /hardware chain inside it that confers some quality of performance that is specific to Naim. Does this also apply to the SSD version and if so that MUST eliminate the HDD inside the US from this argument.
2) If the statements above are correct then by inference the use of a NAS will change the situation. Naim apparently does not say to get the best results with a digital audio chain that includes the US ( or HDX etc) a specific make of NAS hardware is required to preserve the 'Naimness' of rips made on the US. Does this mean that Naim does not include the wider field of NAS hardware in their claim.
3) Does this mean that the ripping process in a Naim server box is necessary to obtain the full benefit of the Naim rip strategy and that therefore systems without a Naim ripper in them are compromised in some fashion.
4) When it is stated the EAC or dB CAN produce bit perfect rips, but only if the hardware there running on allows it, by implication the manufacture of some CD Rom drives is so unusual that the Accurate rip database can be cheated by them and that whole section of the data storage industry happily accepts non-bit perfect archiving of files may occur.
5) How does the Naim label create its download files and are they also going to sound the same as ostensibly the same music files ripped off CDs on a US. For that matter when we download audio files from other respected sources are we missing out again.
I find it hard to believe the above is true. But if it is then it implies the wider world of ripping/storage and transfer of audio is inferior in some way.
Hi Geoff -
There are only a couple of reasons I can think of why WAV and FLAC transcoded on-the-fly to WAV would sound different. Either the transcoder is broken and is delivering a different bit stream, or there is some EMI/RFI proximity effect from the server's extra processing. And yet, both of these seem like very unlikely events. Do you have any theories?
Also, I hope that Guido reads your post. As I recall, the reports that Naim network players sounded best with WAV was one of his frustration points, but it would appear that it is the same with Linn's KDS/1.
ATB.
Hook
No theories I am afraid Hook,
Just a subjective opinion from listening. My audio chain is pretty high resolution and what I am picking up on is a feeling really. The thing I hear particularly is only apparent at higher volume when FLAC seems to take on a slight 'edge' that I don't detect with WAV. This is not a night and day thing. I am quite happy to listen to HiDef downloads which are dominantly FLAC without bothering to do offline conversion to WAV.
Recently I have listened to some ALAC files and have to admit I am not really able to hear any difference compared to WAV but I have only tried a few so far. I will have to dig out some of my favorite silver disks and re-rip them in ALAC with XLD on my MAC to get a better handle on this.
NOTE I am NOT going to re-rip my library. WAV is fine by me.
Sorry I have no logically explanation
Geoff
Hi Geoff
I really am pleased to read your post, as the feeling you describe is exactly as I would describe why I prefer WAV to flac. Haven’t been brave enough to say that here before though. Our system is well below many of you chaps systems but we felt it well worthwhile to have Dbpoweramp convert our flac files to WAV.
Can’t understand why people who have already ripped their CD’s to a lossless format would re-pip to another format when they can just convert them.
Regards
Peter
Hi Dave:
I will try to illuminate each of your excellent points, and thank you for taking the time to answer my many questions - all asked at a beginner's level of understanding in this process. But perhaps this will help others move forward if they want.
- How do you want to rip the Source CDs (dBpoweramp, EAC, iTunes, etc.)
Based on what I have read here, it looks like dBpoweramp is pretty well accepted as a good means by which to do this - so I will load that on my newer laptop - this for ripping my entire CD collection (ugh...) - although the #3 item may introduce a different take on this.
- Which format should you use to store the resulting files (WAV, FLAC, MP3 etc.)
Again based on the things I read here, I am leaning towards WAV, although I may try some experimentation to see how FLAC does as far as a size/quality comparison.
- What sort of storage device should you use (Laptop hard disk, External USB disk, NAS)
Here's where it gets a little interesting - based on your answer, I think I would really like to go NAS; I did a basic google shopping result and found a number of units under $400 US; the question is - what am I looking for? I figured a 4 TB drive should be more than adequate - does it matter if it is SSD vs actual HD? Are there any brands that are preferred or to be avoided? How much memory, and what other features should I get (e.g., # of USB ports? Should I get a CD/DVD drive and put the dBpoweramp on this machine & do my burning here?
- How will you secure backups in case of disaster (2nd NAS, USB Disks)
I would use an external drive of some sort, or even consider burning DVD to allow for "off-site" storage? (Which in the IT departments where I worked was always essential to have something off-site for backup.)
I understand your concern over my current lack of a backup - but at this juncture with the changes upcoming, I am not overly concerned since iTunes is merely being used for my iPhone (and for the time being I can access it from the SB Touch via iPhone) but once I burn my whole CD collection (or even before), this subset will be moved to the new laptop (which already has iTunes) to allow me to place this subset on my iPhone or other Droid later on, for portability in car, gym, etc...with a backup onto my 1 TB external SS drive that I already own. My iTunes music is (and always will be) a subset constrained by the memory available on whatever phone I have. (And as mentioned, I would like to stop doing business with Apple at some point...tired of the way they do many things.)
- What devices do you want to play back your music on? (Main HiFi, Other Rooms, Car etc.)
The rip of the entire collection is for the main system primarily, with the iTunes subset for use as described...and if there is an easy/cheap means by which the second system can "see" the stored music via wireless network, I am open to that as long as it is not especially costly as the second system is just for generally quieter bedroom use in the evening or louder when showering and getting ready...not much "critical" listening on that system.
- How would you like to control your music playback? (Dedicated remote, iPad etc.)
If I can control from my iPhone (now -a s I do with the SBT app) or a Droid later on that would be fine, but not a "deal breaker".
And Geoff - thanks for reminding me that which I already knew if I thought about it that I only need one version of Logitech Media Server - the whole point of which is to see the entire network! (duh!) - based on what I have written above - it likely makes sense to put it on a NAS? Or should I use it on my laptop?
One last question - do the NAS machines work like laptops with regard to being able to be used on US (110V/60 cycle) electricity as well as the rest of the world (220V/50 cycle)? My laptop can be use din either case with nothing more than a plug adapter...do these machines have the same ability?
Thanks to all for putting up with all these questions.
And Geoff - thanks for reminding me that which I already knew if I thought about it that I only need one version of Logitech Media Server - the whole point of which is to see the entire network! (duh!) - based on what I have written above - it likely makes sense to put it on a NAS? Or should I use it on my laptop?
It depends what make NAS you choose. If you google 'Logitech Media Server' and go to the Wikipedia page some way down that page it lists various makes of NAS which LMS will run on and encourages the idea that it is more efficient to run it on the NAS. Basically you keep a NAS running 24/7 which means LMS will always be tracking your NAS Music library whereas if you shut down a PC with LMS on it it will have to refresh its library listing every time you restart the PC.
NAS like PC hardware nowadays is universal voltage so no problems with 110v.
Geoff
Hi Geoff -
There are only a couple of reasons I can think of why WAV and FLAC transcoded on-the-fly to WAV would sound different. Either the transcoder is broken and is delivering a different bit stream, or there is some EMI/RFI proximity effect from the server's extra processing. And yet, both of these seem like very unlikely events. Do you have any theories?
Also, I hope that Guido reads your post. As I recall, the reports that Naim network players sounded best with WAV was one of his frustration points, but it would appear that it is the same with Linn's KDS/1.
ATB.
Hook
No theories I am afraid Hook,
Just a subjective opinion from listening. My audio chain is pretty high resolution and what I am picking up on is a feeling really. The thing I hear particularly is only apparent at higher volume when FLAC seems to take on a slight 'edge' that I don't detect with WAV. This is not a night and day thing. I am quite happy to listen to HiDef downloads which are dominantly FLAC without bothering to do offline conversion to WAV.
Recently I have listened to some ALAC files and have to admit I am not really able to hear any difference compared to WAV but I have only tried a few so far. I will have to dig out some of my favorite silver disks and re-rip them in ALAC with XLD on my MAC to get a better handle on this.
NOTE I am NOT going to re-rip my library. WAV is fine by me.
Sorry I have no logically explanation
Geoff
Hi Geoff
This is veering a bit OT but I find FLACs* to be a bit bright, musically flat / uninteresting and with a slight unreal glossiness to them. On my NDX it is immediately obvious when a FLAC starts playing. I am gradually converting to WAV which is what I find preferable too. I don't find ALACs to be so objectionable but are still not as good as WAV's IMO.
Best wishes, Julian
* uncompressed and compressed, compressed being worse still.
As a KDS/1 owner I can agree that WAV sounds superior to FLAC even when as I do using Asset UPnP to transcode FLAC for delivery to the KDS.
This stands as a separate effect from the ripper used IMO since I am using dBPoweramp rather than a US.
Also more interestingly this holds true for WAV's produced on a PC, a NAS running WHS and RipNas Essenials and a MAC using Patalles to run dB Poweramp in a window all of which I have used to produce my WAV's over time.
Geoff
Hi Geoff -
There are only a couple of reasons I can think of why WAV and FLAC transcoded on-the-fly to WAV would sound different. Either the transcoder is broken and is delivering a different bit stream, or there is some EMI/RFI proximity effect from the server's extra processing. And yet, both of these seem like very unlikely events. Do you have any theories?
Also, I hope that Guido reads your post. As I recall, the reports that Naim network players sounded best with WAV was one of his frustration points, but it would appear that it is the same with Linn's KDS/1.
ATB.
Hook
Hello Hook
Yes I have read Geoff's post and Geoff has been really helpful with advice regarding the KDS vs. NDS. Certainly with my current set-ups Mac Mini into Naim DAC or Vortexbox in to UQ I cannot hear any difference between the compressed and uncompressed formats.
I asked Linn if they would expect me to hear a difference between these formats on the KDS and the answer was no. However, they were not saying that other people would not hear a difference. Just that they would not expect it and the KDS was designed to play all its supported formats with equal aplomb.
My problem with Naim streamers is that Naim appear to advocate strange things such as their streamers sound better with WAV. Perhaps this is true, but not a good message for somebody who avoids WAV because of the tagging problems. Sorry, but I want to use a format that supports tags that iTunes and similar devices can understand - I don't want a Naim only tagging system.
There is also the complexity of keeping the firmware/software up to date on Naim streams - if it requires Windows then that means I'm unhappy with it. It is too expensive to buy the NDS and not be happy with it. There is also my feeling that Naim is proprietary system based on Windows rather than an open system based on Linux.
Naim doesn't support Airplay or any other way of sending the output from my Apple devices to it.
I'm sure the NDS digital out is meant for the Reference DAC 555 which will appear in the not too distant future as a new edition to an ever increasing range of streaming devices and upgrades. I'm confused by the Naim options, whereas Linn's options are easy to understand.
The endless debate on ripping and some Naim staff/dealers people still proclaiming Naim rips better ... when it doesn't - that is not subjective, but a cast iron fact. I think Phil has now confirmed once and for all that Naim creates bit perfect rips that are just like everybody else's.
None of this means that the Naim streamers aren't capable of great sound and bringing immense satisfaction to their users, it just means that I need to look elsewhere on this occasion.
The KDS looks the most promising option for me and I'll audition one after the Olympics; though continuing to use my Mac Mini into the Naim DAC may be all I need to do and upgrade my amplifier. The latest generation of Mac Mini's makes USB/Firewire to S/PDIF convertors unnecessary and a simple optical cable gives superb results. So perhaps an upgrade from the 282 is more in order.
I was very impressed with the NDS/555PS and felt it did surpass what I have - something I've never felt with any CD player. So for the KDS to be worthwhile for me then it has to match or outshine the NDS, as it will undoubtedly be the last streamer I buy in the same way the LP12 is the last record player I'll ever buy. If I thought I'd be able to use the NDS then I'd buy one, but it presents too many complexities for me to ever feel it would be my final digital player (it doesn't seem finished).
All the best, Guy
Hi Guy -
I have always enjoyed reading your posts, and so I wish you well in whatever brand of streamer or dac you choose. Given, as you say, that this will be your final digital music player, then it certainly makes sense to choose the one you feel will work best for you in the long run.
I only highlighted Geoff's post as it seemed to suggest that the KDS/1 was no different from the NDS in terms of favoring WAV over FLAC (at least for some owners), and I had recalled earlier posts from you saying that not having to concern yourself over file formats was an important consideration. Just trying to be helpful, and not in any way suggesting that one is a better player than the other. On that, I have no opinion at this point, but I hope to demo both players by the end of the year.
I think I am in a similar position to you. I, too, am using the DAC/555PS, but am feeding it with an NDX instead of a Mac mini. I do not hear any difference between FLAC and WAV, and I definitely do not hear any difference between WAV and FLAC transcoded on-the-fly to WAV. But I do leave the Asset transcoding option turned on because I have an AVR that shares my UPnP service, and it only plays WAV and MP3 files. Lastly, perhaps I do not hear differences because I use FLAC compression level 0 (minimal compression)? Not sure. Anyway, the only reason I chose FLAC a few years ago was ease of tagging. It appears that WAV has caught up in recent times, and tagging is no longer the big issue it once was.
I am looking forward to reading about your upcoming home auditons. It would be very cool if you could get the NDS and KDS/1 home at the same time for a shoot-out. That should make for a terrific read! I question whether you will find much of a difference between the lossless file formats, but I hope you make this part of your testing criteria, and you let the rest of us know your results. TBH, I doubt you could go wrong with either player, but...
...I also know from experience that there are huge gains to had by going from a 282 to a 552. Were I in your shoes, that would be the first thing I would try as it would benefit both your digital and analog setups.
Good luck with your decisions!
Hook
Guy
With respect, you (& others) are getting far to wrapped up in semantics. Naim supports AIFF and therefore you have a format that also supports iTunes. This thread started out as why Naim prefer the WAV format and later on, it turns out that Linn supports all formats with aplomb. How very unsurprising in the approach from both camps. The thread is turning into a Linn v Naim battle (based on file formats!!!) and IMHO there is no need.
Allen
Allen
Wit respect since we are apparently getting wrapped up semantics lets get the semantics correct. The thread did not start out on 'why Naim prefer WAV' It was prompted by claims that Naim hardware produces rips that are better than ostensibly the same rips done by other routes, and it was Naim personnel that introduced the fact that WAV sounds better than FLAC in their opinion, to the ensuing discussion. Nobody that I am aware of has made any statements that could be construed as suggesting they are taking part in a Naim vs Linn battle.
After all both suppliers make excellent streamer DACs and it is up to the individual to decide which product is most to his or her liking. Guy just did that, explaining his position comprehensively in the process but I don't see that is any different from Naim expressing their opinions earlier in the discussion.
Geoff
Hi Geoff
This is veering a bit OT but I find FLACs* to be a bit bright, musically flat / uninteresting and with a slight unreal glossiness to them. On my NDX it is immediately obvious when a FLAC starts playing. I am gradually converting to WAV which is what I find preferable too. I don't find ALACs to be so objectionable but are still not as good as WAV's IMO.
Best wishes, Julian
* uncompressed and compressed, compressed being worse still.
Hi Julian
Thanks for the input, It will be interesting to see how I get on with comparing WAV and ALAC.
Anyway, the only reason I chose FLAC a few years ago was ease of tagging. It appears that WAV has caught up in recent times, and tagging is no longer the big issue it once was.
Hook
Yes I stumbled into this and by luck started ripping to WAV with dB Poweramp which does WAV tagging without realising that was unusual and was puzzled when I read that WAV didn't support tagging. Lucky me.
regards
Geoff
Geoff, just to add, as there seems to be a huge misunderstanding based on perhaps partial support of WAV files by some consumer software like iTunes and others. The WAV format has always supported metdata. This standard format is more suited to industry rather than consumer and therefore it doesn't support 'album art' as well as custom tags about ripping etc and these additional custom tags are not garenteed to hold the same meaning between applications.
Then recently certain vendors unoffiicially bolted the mp3 tagging method based on ID3 onto WAV (which is entirely legit due to WAVs open RIFF format). It is this id3 data that causes the confusion as it is not an 'official' part of the WAV specification, and is not supported by all software that fully supports WAV metadata,
Software such as DBPoweamp, by default, writes both the normal WAV tags as well as the unofficial ID3 tag for interoperability with other software.
An example, if you don't have dBpoweramp installed on a windows OS if you right click on a WAV file you will see the standard WAV metadata. Install dBpoweramp, do the same, and you see a new ID3 info screen which additionally contains clip art and extended tags like the Accuraterip confidence value, ripping history etc.
It is this flexibility that IMO makes a WAV a superb format for ripping, it evolves with new requirements/ developments without having to re rip your entire CD collection. This is not open to you with FLAC. I have no idea whether AIF allows custom chunks to be added, even though it came from the same stable as WAV (RIFF) all those years ago.
Simon
Hello Simon - I agree with what your are saying, but being an Apple/Amiga user dBPA is not an option for me (I'd use it if ran on OSX or Linux) - so my requirement is to use an Apple friendly format - I also like to embed the artwork - just my way. Unfortunately, the Naim engines store metadata and artwork in a way that I don't find easy to use. Yes I could spend lots of time sorting it out, but it appears the KDS and some much more modest streamers just work (in the way that suits me). What I want is a device that doesn't require me to worry about what format the file is in, it just plays it and makes any desire for me to convert from one to another irrelevant and I don't feel the NDS does this. So it is really me being me and just saying if I'm paying lots of money for it then this is what I want it to do. WAV does seem to be more for Windows users than non-Windows users in terms of utilities that support it.
Hello Allen - My argument isn't really a Linn vs. Naim one ... simply responding and trying to explain why I'm not a potential customer for the NDS. Great box, just not for me. My biggest issue is the software/firmware update process, which is for Windows users only. I wouldn't buy any product where this was the case; I didn't know it was the case when I bought the UQ and was caught out when it couldn't be updated by simply selecting update on the menu.
Hello Hook - I'm unlikely to do a shoot out at home (possibly at my dealers). The audition is really to ascertain if I like the KDS enough to buy it. I have no qualms that the NDS when set up to sound at its best is quite superb, but unfortunately that doesn't matter, as there are too many usability points for me (and may be only for me) that make it unsuitable. I, too, enjoy reading your posts, which are always helpful and good humoured and, in common, with Geoff's, Simon's and Allen's and many always interesting to read through.
All the best, Guy
Hi Allen
I Appreciate where you are coming from.
I agree the first and most important requirement is that the audio performance from a streamer/DAC be what someone is looking for and that the 'bang for buck' factor makes sense but it doesn't end there. Given very much the same standard of performance from two different manufacturers secondary considerations come into play. For example is a multi box solution ergonomically acceptable vs a single box. Does a solution that offers to take care of a lot of the elements involved in ripping, cataloguing and playing suit over a solution that requires the user to set these up outside the supplied system....and so on.
Us folks on here who happily delve into the subtleties of ripping strategies, NAS setup, UPnP severs and control points and network connectivity etc are probably viewed as 'nerds' by the vast majority so we reflect a pretty small section of the wider community who don't want to get involved and 'just want it to work' That why folk like Sonos and Meridian at different price points are attractive.
Geoff
Interesting post Geoff - Linn don't seem to consider this an issue but Naim do. Some interesting threads on the Hi-Fi Critic forum about CD ripping and differing sound quality, even if the checksums match,,,.
My XLD test rips of a Naim label CD vs the same tracks as downloads from the Naim label sounded the same so i've not worried about it.
Keeping an open mind on this though.
Perhaps Linn don't make an issue of it because they have avoided making their own ripping solution? (It would have been much easier for us to have gone that route but we have a "source first" ethos and so we wanted to make sure our source material was first-rate.)
One of our bigger dealers are also definitely Linn-centric as far as their staff are concerned (4 of their 5 staff are dedicated Linn-ites) but all of them use UnitiServes as sources for their home systems because they believe that they sound better than their existing ripping solutions - one of them had already ripped his 5,000 CD collection as FLACs (as he had been advised to by Linn) and yet openly admitted to me that he was re-ripping his entire collection using a UnitiServe SSD as it sounded better.
Phil
Dont know if its already been said but Linn do have a ripping solution its called the Linn Kivor, either the index or the tunboks. Both have CD drives that are used to rip CDs to local hard disks to be played back locally via a built in dac on the index or streamed from the tunboks.
They actually rip CDs to WAV format.
Interesting post Geoff - Linn don't seem to consider this an issue but Naim do. Some interesting threads on the Hi-Fi Critic forum about CD ripping and differing sound quality, even if the checksums match,,,.
My XLD test rips of a Naim label CD vs the same tracks as downloads from the Naim label sounded the same so i've not worried about it.
Keeping an open mind on this though.
Perhaps Linn don't make an issue of it because they have avoided making their own ripping solution? (It would have been much easier for us to have gone that route but we have a "source first" ethos and so we wanted to make sure our source material was first-rate.)
One of our bigger dealers are also definitely Linn-centric as far as their staff are concerned (4 of their 5 staff are dedicated Linn-ites) but all of them use UnitiServes as sources for their home systems because they believe that they sound better than their existing ripping solutions - one of them had already ripped his 5,000 CD collection as FLACs (as he had been advised to by Linn) and yet openly admitted to me that he was re-ripping his entire collection using a UnitiServe SSD as it sounded better.
Phil
Dont know if its already been said but Linn do have a ripping solution its called the Linn Kivor, either the index or the tunboks. Both have CD drives that are used to rip CDs to local hard disks to be played back locally via a built in dac on the index or streamed from the tunboks.
They actually rip CDs to WAV format.
Do they still make the Kivor. I though that was part of e old Knekt system pre DS ?
James
PS - Geoff see FB re XLD :-)
Phil, re reading your post.. These linn-ites you refer to, are they using the unitiserve as the upnp / DLNA server? In which case it is possible it sounds better because of the TCP window sizes used by the unitiserve expand to a larger value than their other upnp server. I have mentioned before, window sizes can effect audio quality on the NDX into NDAC / 555PS and so it is possible this is what they are hearing on their linn network player. This is of course nothing to do with the rip if this is the case, but the upnp server TCPIP stack.
Larger window sizes cause less processing on the TCP processor on the network player. If this processing causes noise, then by doing less of it you are producing less noise. Of course Ethernet line level signals have the propensity of generating a lot of noise on the Network streamer electronics that needs careful decoupling and screening. And Iguess there is always leakage here.
It was quite a revaltion when a certain dealer on this forum and I discovered this when we were experimenting with different upnp servers and how they sounded playing the same NAS files.
Simon
Hi Simon -
Tuning TCP window size has been on my to-do list for at least a year -- ever since you first raised the topic! Am going to give it a go when I get home from vacation. Sounds like the only place I need to worry about is my W7 PC running Asset...
Was just doing a bit of reading, and saw that W7 has TCP auto-tuning turned on by default, and if I am reading these tech notes correctly, this automatically sets window size based on traffic levels. So it sounds like step one is turning this off?
IIRC, you are using WHS? Did you have to turn off its auto-tuning? And did you set window size to the max (65535) number of bytes?
Will coninue to research, but would appreciate a pointer to a step-by-step if you happen to know of one, as well as recommended window sizes to try. Is it as simple as editing a registry parameter and rebooting?
Thanks a lot!
Hook
Good post Hook, I can imagine many of use would be interested if Simon can offer any further information on this.
Peter
HELP!!!
OK guys no clue about this so copied from the WiKi page:
"The throughput of a communication is limited by two windows: the congestion window and the receive window. The former tries not to exceed the capacity of the network (congestion control) and the latter tries not to exceed the capacity of the receiver to process data (flow control). The receiver may be overwhelmed by data if for example it is very busy.
Also because many routers and firewalls do not properly implement TCP Window Scaling, it can cause a user's Internet connection to malfunction intermittently for a few minutes, then appear to start working again for no reason.
There is also an issue if a firewall doesn't support the TCP extensions."
Excuse me if I am being dumb here but In particular I am interested by the bits I have highlighted in red.
Correct me if I am wrong but the receive window is up to the streamer to provide and since streamers have one specific data transfer task they are pretty crude computers right?...and I assume are most likely to have a fixed receive window size which you can't do anything about. So that must limit what you can do to the Congestion window anyway.
The second point about routers would also seem to be pretty relevant as far as I can see.
Sooo.. is there really much you can do above and beyond what is already set with a typical Router/NAS/ streamer DAC setup?
regards
Geoff
Hi Hook
Ok, this is an area I am experimenting with.
http://support.microsoft.com/kb/224829
I am looking to set the scale factor progressively higher on my upnp server TCP kernel until I hear no change. The scaling factor is used to set the max window size in multiples of 64 kbytes. It might be however the NDX can only support 64kbytes in which case you would ideally set that as the send window size.
I use Windows Home Server v1 for my upnp which is based on windows server 2003. Hence the post with registry settings to modify.
If you use Wireshark you can look at the Nacks and see the largest window size the TCIP protocol machine / kernel can support in the transport session between upnp and NDX.
I am sure whether explicitly or implicitly this has been optimised by Naim between the unitiserve and NDX/NDS TCP kernels.
Simon
Hook
I have just set to the TcpWindowsSize to 2097120 dec.
This is in HKEY_LOCAL_MACHINE\System\CurrentControlSet\Services\Tcpip\Parameters
I have created the DWORD value TcpWindowSize and entered the value 2097120 dec then restarted the machine (WHS).
This overrides the autosizing alogorithm and sets the send max windows size to 2M - subject to the NDX. I will be listening shortly to the differneces - but my listening room is in use right now :-o
Simon
Thanks a lot Simon!
Am pretty beat after a long work day, so will probably wait till the morning before trying this. Thanks a lot for the directions -- am looking forward to giving this a try, and will report back on what I hear!
Hook
I decided to do a little experiment today, because my sonos doesn't support 24 bit I decided to setup an old acer revo I've got and use the optical spdif to play 24/192 through my dac. I found that unfortunately the spdif only supports a max of 24/96 but still plenty.
So with an acer revo connected to a rega dac via optical spdif and using the J River media player set to kernel streaming 24/96 definitely sounded better than 16/44 on the same song. So much so I'm considering ditching the sonos and making the PC permanent. Also the 24/96 vinyl rips I tried sounded brilliant as well.
In my opinion the 24 bit sampling depth makes one hell of a difference.
I decided to do a little experiment today, because my sonos doesn't support 24 bit I decided to setup an old acer revo I've got and use the optical spdif to play 24/192 through my dac. I found that unfortunately the spdif only supports a max of 24/96 but still plenty.
So with an acer revo connected to a rega dac via optical spdif and using the J River media player set to kernel streaming 24/96 definitely sounded better than 16/44 on the same song. So much so I'm considering ditching the sonos and making the PC permanent. Also the 24/96 vinyl rips I tried sounded brilliant as well.
In my opinion the 24 bit sampling depth makes one hell of a difference.
Dude
Just to clarify....
Did you run both the upsampled rip and the original 16/44 file via the Acer?
Or one via the Acer and the other via the Sonos?
Both files via the acer, straight comparison.
The original song was a CD rip?
Upsampling can appear to make things sounds smoothed out...
By rights, 24/88.2 should be better than 24/96 (assuming the file is a CD rip).