Magellan

Just a correction to my own post. I just remembered Roon can do resampling and some other signal processing. I have not seen any actual measurements comparing the signal processing in Roon with other software. For the time being I would not use Roon for signal processing.

As for sound quality there is no differences between different PC or MAC software players - as long as the computer or software is set up not to alter the bits. That is, as long as the data is bit perfect transported to a external DAC with proper jitter correction, then sound quality is equal between the mentioned players. If re-sampling or any other form of signal treatment is wanted, in Roon it has to fall back on the operating systems own audio engine. If re-sampling is needed fore example, I would strongly recommend a MAC and not a PC, if Roon is the choice. Windows audio engine is rather poor compared to OSX. If Windows is the preferred system, and If resampling or any other sound processing is wanted or needed, then Jriver (or foobar2000) is the obvious choice. It´s inbuilt audio engine is superb compared to others. I would say no other software-player comes close. I agree that Roon has a very nice TIDAL integration, but that is also in my opinion it´s only outstanding strength. In my view, the big drawback is the need for a traditional PC or MAC. Roon is a relatively demanding system on resources. You can for example not use an Android smartphone as remote - as you can with most other software players - but you need a rather powerful tablet. If you want to use an iPad, you need a rather new model. etc.. Me myself have become tired of powering up a traditional computer every time I want to listen to music. I have therefore built a small RapsberyPi+HiFi-Berry computer, and as a player I use Volumio and MPD. Since this system is DNLA compliant, and as I have al my music on a NAS, it is very convenient. AND - the great thing is - Jriver is also DNLA compliant, which mean I can remotely control the Volumio player with Jriver on an laptop. If I want to read about the music I am playing, like in Roon, observe then that Jriver is linked to Wikipedia (and Google if wanted) and can provide tons of info within the system. I find this DNLA feature in Jriver outstanding. If I don´t want to fire up a laptop and Jriver, then I use BubbleUpnp on my Android smartphone instead, and control the Volumio player in that way.

"Definitely more affordable than the Dirac" How do you know?

Who needs 192 kHz?

I have been using this software for a couple of years now. First, the effect depends upon the room in which it is used, i e the more "difficult" the room, the more pronounced the effect. However, of all audio tweaks I have tried during the years, this is overwhelming in it´s effect compared to most other tweaks. Dirac make a huge difference compared to the more subtle changes resulting from the more "ordinary tweaks". But, you will not be able to utilize the full potential of the software if you do not get an accurate measurement. Therefore, a calibrated microphone is a must, unless the microphone used are not one of the few very transparent and very expensive microphones. Regarding the price of the software, if it is expensive or not, I would say price is something very relative. But as far as it has to do with value it has to be compared to other investments in audio equipment. As many audiophiles don´t hesitate to put hundreds of dollars into DAC:S, computers etc.. I would say, compared to most other investments, as long as the hardware are of decent quality, in most cases I think Dirac is one of the most sound improving investment possible.

It seems this was the problem. Since I marked the Synology for auto-mount the problem with skipping files has disappeared. Thanks for the help.

True, there is not a perfect filter. Well, I can buy this. I will do some experimentation during the next couples of days. Tis is also in line with superdads observations I think. Thank you superdad for sharing your thoughts and observations.

My DAC is not the same as in the TNT-audio review. I guess it is this review you read. My DAC is the more advanced pre 906. But, it utilize the same CS4398 DAC-chip, so your argument is valid anyway. Yes, I understand how you mean. The oversampling could potentially be different or better done by a software. I will try this out on my new MAC with Audirvana. My understanding is that the oversampling done in the DAC is done because it makes it possible to avoid a very step filter. It makes it easy to place he filter in a frequency above the human hearing. And I have always thought the oversampling in the DAC is just about adding a lot of zeros and not adding anything "substantial" to the "original" file. But my understanding regarding this is limited. Another new interesting question will rise - I think - if we oversample in real time in Audirvana. If oversampling is done, then CPU-overhead will no longer be minimized. And oversampling is, what I understand, always done in floating point mode. So the potential benefit with integer mode is lost as soon as oversampling takes place. Am I right? Thank you for your "DAC-research". I am sorry if I could be read as brusque in my previous posts. It was not my intention.

It was my intention to leave this discussion now since it by now must be rather meaningless for most other participants in this forum. But since Jud writes this seems to him to be an unsatisfactory way to leave things, I think I have some kind of obligation to answer his latest post and explain how I meant in my posts and how I perceived Juds posts. Yes I wrote “Believing is believing and a fact is a fact”. But in a post before that I also wrote “I read this paper, but it didn´t make me feel enlightened. Besides, the paper claims things which are - for what I know - not proven, just based on assumptions.” 1. What I meant was, I have not read any article presenting any proof based upon facts that show the existence of CPU-load induced jitter. Those articles I have read, among them the article I linked to, indicate the opposite. 2. My point when writing about facts and believes is that the measurements brought up in the linked article are facts. The method behind the measurements can be bad or biased and not relevant or misleading. In other words, the data can be correct, but without relevance due to the method used. In the same post where I wrote believing is believing …etc.. I also wrote: “if these measurements are not correct or relevant I would be happy to know Why”. Meaning: I am open to any explanation based upon facts that falsify the conclusion drawn in the linked article. After I wrote this Jud commented a third persons post with: “Indeed, I thought I was avoiding this sort of thing by advising Magellan to be happy and content as he was.” It was this specific post I commented with “ my initial question was just about the integer mode, ie how it works and what it does. It seems that asking for a rational explanation concerning this is very no no..” From this comment I then was accused to dismiss the paper written by Audirvanas originator as non rational. Well I didn´t. I just wrote I didn´t feel enlightened by the article. And then I wrote that the paper claims things which are - for what I know - not proven, just based on assumptions. And an assumption is not a fact. I could write a lengthy article about the paper and what I find being vague in it, but that is another story. But I did not say the article was not rational. Given the assumptions are correct, I think it is very rational and reasonable. But, as I have said, I have some doubts about these assumptions. I can mention that I have done extensive blind testing myself and with audiophile friends, and when blind testing none of us have been able to discriminate between music played under heavy CPU-load and with very low CPU-load. I have also been experimenting a lot with very low power minimalistic computers, fed by clean linear power supplies, running very bare bone Linux operating systems, no graphical overhead etc.. And truth is, I can´t discriminate between these computers and – for example, the Mac mini running a normal OSX setup. The reason I did not answer the question about the DAC is that it is not one of the major international brands. It is built by a small custom-building company. www.harmonydesign.se

In general I think it s nice when people read what Is actually written and not attribute me arguments or conclusions which I have not written myself. And I think some people here has a rather ignorant and arrogant tone, not polite at all. Thats all. Have a nice day everybody..

Well, my initial question was just about the integer mode, ie how it works and what it does. It seems that asking for a rational explanation concerning this is very no no.. Be happy and keep quiet. Then you are suddenly in a subjectivist - objectivist war. Tiresome.

Believing is believing and a fact is a fact. For those who also like some facts and not just assumtions and believing, here is a link to meassurements where CPU-load induced jitter is the issue. These meassured results are facts, not believing, and if these meassurements are not correct or relevant I would be happy to know Why. http://archimago.blogspot.ca/2013/03/measurements-hunt-for-load-induced.html I would have been happy and I would just go on with Audirvana since I like the software - if it was not for the iTunes & Audirvana combo skipping through the files like i describe in my posts above.

No, this could not be true. iTunes is just used as a graphical interface. All the actual playing of the files is handled by Audirvana. I can´t hear any difference at all.

Inside my DAC..hmm..yes, but in what way has what happens inside my DAC to do with what a software music player does? Yes, I read this paper, but it didn´t make me feel enlightened. Besides, the paper claims things which are - for what I know - not proven, just based on assumptions. Just an example, it is often said higher CPU load degrades sound. Actual measurements, where data has been compared, show no impact from cpu-load, as long as the cpu-load is what could be called normal.

I have used JRiver on the Windows platform and know the program well. But I wonder what this integer mode means. A music file is either played bit-perfect or it isn´t. If there is no upp- or downsampling involved, then if two players really sound different then at least one of them is not bit perfect. Is integer mode necessary for bit-perfect on the MAC-plattform? If so, I suppose players without integer mode can´t play bit perfect-