jkeny wrote:Ok, this is getting "fuzzy"
- Guess each hardware differences between yours & Pearse's has to be identified & tested?
- same network driver
- different DAC Ciúnas - but this has worked at US=30 on Ken's system
It will be interesting to get to the bottom of this
my ethernet adapter chipset is Realtek RTL 8168 same as frizzup on jplay fourm with the tplink adapter
we got us=30 (never thought of trying for less) on single mode at kens, we only got 180 with streaming via wifi no crossover cable
if pearses can do it so can frans we just have to figure it out
sd card player, modded soekris dac, class a lifepo4 amp or gb class a/b amp, diy open baffle speakers based on project audio mundorf trio 10's
US works fine on franzuma standalone
xtream works AOK in streaming mode
same zuma driver as pearse
Few things yet to try incl turning off FW and defender (I do have the 48168 port open on the FW though)
Power management: set to best performance but if there are any fancy settings in the device manager for the network let me know!!
have yet to run performance monitor and set us running and see what the CPU demand goes to
Thats the same network hardware tried with 2 different drivers and no change. What else in there could be stopping this? Seeing as nothing has changed on the control PC, it kinda must be something to do with the franzuma?
US works fine on franzuma standalone
xtream works AOK in streaming mode
same zuma driver as pearse
Few things yet to try incl turning off FW and defender (I do have the 48168 port open on the FW though)
Power management: set to best performance but if there are any fancy settings in the device manager for the network let me know!!
have yet to run performance monitor and set us running and see what the CPU demand goes to
Thats the same network hardware tried with 2 different drivers and no change. What else in there could be stopping this? Seeing as nothing has changed on the control PC, it kinda must be something to do with the franzuma?
Fran
open a homegroup between pcs? maybe
out of ideas for now
sd card player, modded soekris dac, class a lifepo4 amp or gb class a/b amp, diy open baffle speakers based on project audio mundorf trio 10's
Ultimately this seems a technical issue with how ultrastream is handling network drivers and it will require Josef to recognise that maybe some code change is required in US to get over this. I am surprised it hasn't had a larger volume of complaints. It seems very strange that a later version of a driver for a network card that did work with US now fails to work after update. It kind of dispels the view Josef has that the network card manufacturers haven't updated their code for windows 8 and it is their problem.
I wonder is it mainly older network cards that seem to be fine with US and not the later versions? Ken and Fran both have old pcs. I don't think Johns laptops are in the first flush of youth. Maybe Josef developed US on an old PC/Laptop. It is a pity because it puts jplay back into the space where some punters will be nervous of taking the leap especially when they follow a thread like this.
tony wrote:Ultimately this seems a technical issue with how ultrastream is handling network drivers and it will require Josef to recognise that maybe some code change is required in US to get over this. I am surprised it hasn't had a larger volume of complaints. It seems very strange that a later version of a driver for a network card that did work with US now fails to work after update. It kind of dispels the view Josef has that the network card manufacturers haven't updated their code for windows 8 and it is their problem.
I wonder is it mainly older network cards that seem to be fine with US and not the later versions? Ken and Fran both have old pcs. I don't think Johns laptops are in the first flush of youth. Maybe Josef developed US on an old PC/Laptop. It is a pity because it puts jplay back into the space where some punters will be nervous of taking the leap especially when they follow a thread like this.
Im fairly sure all mine work there fresh enough
I dont think its an ethernet problem this time
sd card player, modded soekris dac, class a lifepo4 amp or gb class a/b amp, diy open baffle speakers based on project audio mundorf trio 10's
I puts me money on memory management being the issue. I think when he moved away from grabbing a chunk of contiguous memory for Jplay's exclusive use, he entered into a more dangerous memory management field with possibly more dependencies than he is aware of or is able to control.
It would explain why the results are variable & fuzzy as we are finding & Peter_T posted on Jplay forum. I believe this is a revealing post (& worth copying below) & should be looked upon by Josef as good feedback but his head doesn't seem to be at that place at the moment. He needs to borrow a system that has these issues as according to him, his system seems to be working at US=20 with no optimisation or attention!
It appears to be difficult to measure the lowest US : it is dependent on whether you test it directly after reset, or whether you test it after you have closed and opened Jmini a few times or whether you use a short or a long track.
Example in my system. Single AudioPC (Shuttle with Atom525, 4 GB) Jmini, Minicache=250, XS=500, DirectLink, native bitstream, Ustream, driverbuffer (DB) = 0. DAC-driver is M2tech driver for Minimax-plus (same driver as for M2tech Young DAC).
US=60. Fresh restart. Will play at some startups, not at others. US=70 fresh restart, will play small tracks (1-1.5 minutes, not longer ones). US=100 will reliably play all tracks. If I lower US (without computer reset) in steps till there is no sound, then I would expect that sound would come back with the next higher step in US. But no, I have to increase to several steps higher than before. Probably because closing/opening Jmini changes the memory fragmentation?
So if you found an "absolute minimum" US value, increase it by 30% in order to get into a workable zone. (someone has said that before on this forum).
Ustream (single PC) is not dependent on an ethernet card. I switched it off and Jplay could play local files.
Going up from Direct-Link to 2 gives no sound (US=100, DB=0).
You need to increase US for higher DB's : DB=8 with a minimum of US=4500 in my system. In this situtation I can increase DirectLink to 32 before sounds stops.
I tested these values in 3 different settings. AudioPC started with regular desktop W8 from SSD, AudioPC started with John Doe's W8 on an USB stick (lots of services less than normal W8) to the desktop interface or to the Total Commander interface. This did not bring any significant changes in the US values mentioned above
Lesson : parameter settings are a bit "fuzzy" even within one system. Strict rules are therefore difficult to give.
www.Ciunas.biz
For Digital Audio playback that delivers WHERE the performers are on stage but more importantly WHY they are there.
[quote="jkeny"]I puts me money on memory management being the issue. I think when he moved away from grabbing a chunk of contiguous memory for Jplay's exclusive use, he entered into a more dangerous memory management field with possibly more dependencies than he is aware of or is able to control.
The one difference I suspect between Pearse's zuma and Fran is the amount of ram. 16gb as opposed to 8gb. Is that a solution? or is the problem deeper than that. But that doesn't explain older pc's with lower amounts of ram working fine.
FWIW,
my AudioPC is about 2 yrs old and uses an Intel 82567V-2 Network Adapter Driver 10.1.17.0 7Gb RAM and
my StreamerPC is ancient and uses a Netgear WNDA3100v2 N600 Wireless Dual Band USB adapter and 4 Gb RAM. The original Broadcom 440x 10/100 Controller didn't like Ultra and is disabled.
KM
