Sure. But as we all know, that doesn't necessarily mean there is an actual difference.
Yes perhaps, strangely in my forum same like more the newest and same the earlier.
Inviato dal mio SM-N9005 utilizzando Tapatalk
And in a proper blind test neither group would be able to tell any difference, by far the most likely outcome.
You can think all that you want
. , simply in my set up the differenze is clear, and after I returned at the earlier version my ears gone happy..Ciao
Inviato dal mio SM-N9005 utilizzando Tapatalk
My actual settings brings very pleasant results and responsiveness:
An externally hosted image should be here but it was not working when we last tested it.
I hear absolutely zero difference between this version and the last version (both sound as good as they can on my setup).
Those buffer settings shouldn't do anything at all to sound quality.
Basically:
- the stream buffer determines the amount of "hiccups" in a network the playback can withstand. Given a 320kbps stream, i.e. 40KB/s, every 40KB of buffer size will roughly give you one second of bridging any interruption in the network. So with 512KB, you can pull the network cord for a bit more than 10 seconds and don't notice any drops
- the output buffer determines the amount of "hiccups" the system itself can stand, i.e. a sound card that's not ready to receive new data, or some process taking up resources preventing squeezelite from accepting packets.
So basically: for crappy streams/slow networks, you might benefit from a slightly higher stream buffer. For a busy system, you might benefit from a slightly higher output buffer. But in general, if the defaults aren't ok for you, something else is wrong in your setup and fixing that should have priority (i.e. fix the network, don't run other stuff on your system).
Also note that an undersized buffer doesn't cause any problems with the frequency spectrum, i.e. you won't get less bass or less highs from these settings, just hard dropouts if you set them too tightly.
Look at the sound card like taking a scoop of water out of a bucket every second/millisecond/whatever.
Changing the size of the bucket won't change the amount of scoops that the audio chain takes out of it, because there's a fixed amount of samples per time period in the file. It'll just change the amount of water you have before the bucket runs out. In a good setup, the bucket just never runs out of water.
Those buffer settings shouldn't do anything at all to sound quality.
Basically:
- the stream buffer determines the amount of "hiccups" in a network the playback can withstand. Given a 320kbps stream, i.e. 40KB/s, every 40KB of buffer size will roughly give you one second of bridging any interruption in the network. So with 512KB, you can pull the network cord for a bit more than 10 seconds and don't notice any drops
- the output buffer determines the amount of "hiccups" the system itself can stand, i.e. a sound card that's not ready to receive new data, or some process taking up resources preventing squeezelite from accepting packets.
So basically: for crappy streams/slow networks, you might benefit from a slightly higher stream buffer. For a busy system, you might benefit from a slightly higher output buffer. But in general, if the defaults aren't ok for you, something else is wrong in your setup and fixing that should have priority (i.e. fix the network, don't run other stuff on your system).
Also note that an undersized buffer doesn't cause any problems with the frequency spectrum, i.e. you won't get less bass or less highs from these settings, just hard dropouts if you set them too tightly.
Look at the sound card like taking a scoop of water out of a bucket every second/millisecond/whatever.
Changing the size of the bucket won't change the amount of scoops that the audio chain takes out of it, because there's a fixed amount of samples per time period in the file. It'll just change the amount of water you have before the bucket runs out. In a good setup, the bucket just never runs out of water.
If you're running Daphile on a reasonably modern PC with 4 gigabytes of memory, then setting the output buffer large - say 1 gigabyte - then a whole track will be loaded into that output buffer at the start and there won't be any need to access the disk or network during play - roughly equivalent to the memory play facility in JRiver or Audirvana.
Now, whether this makes any difference or not is an interesting question, but given the basic idea of Daphile is to do the minimum necessary to play music, it's a sensible option in line with that philosophy.
Alan
Now, whether this makes any difference or not is an interesting question, but given the basic idea of Daphile is to do the minimum necessary to play music, it's a sensible option in line with that philosophy.
Alan
Also I think as Keres: The latest version does not fully satisfy me like the previous one, especially in terms of dynamics and bass register. The previous version has a full sound, dynamic, powerful and deep bass register, and maintain a detail completely right.You can think all that you want
Ciao
I have conducted tests to compare very careful and repeated the conclusion is what I have already said.
I listen daphile with a thin client VIA 7 (mini PC), USB-I2SoverUSB transport, a DAC with 6xPCM1704 and output I / V conversion with ECC99 and ECC88 tubes, Lightspeed Attenuator and Amplifier My_Ref Fremen Edition , speakers Diapason Adamantes II
Last edited:
Sontero is right... The previous version was much better in dynamics and bass. Bass is a little boomy now, te previous version was much more detailed and musical.
Same version, but 64 bit sounds best for me
I am extremely annoyed by these claims in sound difference, as they seem to be completely unsubstantiated and based on unreliable "tests". The reason why I'm particularly annoyed, is because this may influence new and existing users to use older versions, giving them an otherwise degraded experience because they're not able to use fixes and new features, plus it increases the burden on the maker because he will get lots of complaints from older versions that may have long been fixed. It's just confusing and annoying to deal with, for all of us.
Now, all of this would have been different if there was real evidence for the difference in sound, in the form of at least a plausible explanation in terms of changes in settings or software, or a real measurement between versions. Something like RMAA would be ideally suited for measurements, it's not hard to do, it just requires a decent sound card to do the measurements. Considering the amount of money people with these complaints are willing to spend on sound quality, it almost seems silly that hardly any of this budget is spent on actually measuring real results.
Now, all of this would have been different if there was real evidence for the difference in sound, in the form of at least a plausible explanation in terms of changes in settings or software, or a real measurement between versions. Something like RMAA would be ideally suited for measurements, it's not hard to do, it just requires a decent sound card to do the measurements. Considering the amount of money people with these complaints are willing to spend on sound quality, it almost seems silly that hardly any of this budget is spent on actually measuring real results.
This is why the strict rules of the hydrogenaudio forum were established. Real results - properly functioning audio software packages - grow from such environment free of voodoo and feelings.
This is not to say that diyaudio does not breed properly functioning software - daphile being an example. But I understand why development threads of so many core audio technologies prefer to be located there, shielded from the tsunami of never proven audiophile claims based on feelings only.
This is not to say that diyaudio does not breed properly functioning software - daphile being an example. But I understand why development threads of so many core audio technologies prefer to be located there, shielded from the tsunami of never proven audiophile claims based on feelings only.