Topic: NEW: Radio Paradise (Read 58017 times)

bob · « **Reply #150 on:** November 16, 2018, 03:33:34 pm »

We've added the other Radio Paradise channels and that will be in an upcoming build as well.

Solo-53 · « **Reply #151 on:** November 17, 2018, 10:45:10 am »

Quote from: bob on November 16, 2018, 03:33:34 pm

We've added the other Radio Paradise channels and that will be in an upcoming build as well.

Thank you very much for this news!...I've been waiting for the division of RadioParadise streams by genre preferences...

newsposter · « **Reply #152 on:** January 04, 2019, 12:24:22 pm »

This weeks adds/changes on RP are streaming just fine for me!

jac61gta · « **Reply #153 on:** January 21, 2019, 12:30:55 pm »

Hi, just upgraded to 24.74 and I'm having the same problem others were having with regards to the RP song streaming then cutting out ~5-10 seconds before end, silence for 3-10 seconds then plays again for a few seconds, then the next song starts. This happens with every song, while the next song loads. I've read through the 4 pages here and see others have had success changing some parameters. My parameters are the same as those that worked for others, but no joy. I'm using DLNA to a LUMIN and also WASAPI direct to DAC (usb) WIN10.

Memory playback turned off
Gapless
Do not play silence - unchecked
Use gapless for sequential album tracks - checked
Use gapless for manual track changes - checked

Any other suggestions are greatly appreciated.
Many thanks.

bob · « **Reply #154 on:** January 21, 2019, 02:01:30 pm »

Quote from: jac61gta on January 21, 2019, 12:30:55 pm

Hi, just upgraded to 24.72 and I'm having the same problem others were having with regards to the RP song streaming then cutting out ~5-10 seconds before end, silence for 3-10 seconds then plays again for a few seconds, then the next song starts. This happens with every song, while the next song loads. I've read through the 4 pages here and see others have had success changes some parameters. My parameters are the same as those that worked for others, but no joy. I'm using DLNA to a LUMIN.

Memory playback turned off
Gapless
Do not play silence - unchecked
Use gapless for sequential album tracks - checked
Use gapless for manual track changes - checked

Any other suggestions are greatly appreciated.
Many thanks.

What are your DLNA server settings? Can your Lumin play flac natively?

jac61gta · « **Reply #155 on:** January 21, 2019, 02:12:51 pm »

Hi Bob, thanks for the reply.
Yes LUMIN plays FLAC. In fact JRiver streams the RP FLAC with no problem over DLNA and sounds great! The problem described above (and I believe user UncleMilt had the same issue) occurs using DLNA or WASAPI direct (USB) to my DAC. My download and upload speeds are 32mb and 11mb so no problem there...
DLNA - Audiophile 24 bit
Audio:
Specified output
Mode - PCM 24 bit
No DSP
DLNA Server, Renderer, Controller all checked.
Thanks,

bob · « **Reply #156 on:** January 21, 2019, 03:43:46 pm »

Quote from: jac61gta on January 21, 2019, 02:12:51 pm

Hi Bob, thanks for the reply.
Yes LUMIN plays FLAC. In fact JRiver streams the RP FLAC with no problem over DLNA and sounds great! The problem described above (and I believe user UncleMilt had the same issue) occurs using DLNA or WASAPI direct (USB) to my DAC. My download and upload speeds are 32mb and 11mb so no problem there...
DLNA - Audiophile 24 bit
Audio:
Specified output
Mode - PCM 24 bit
No DSP
DLNA Server, Renderer, Controller all checked.
Thanks,

You get the break when playing directly to a USB connected DAC??

Shady Bimmer · « **Reply #157 on:** January 21, 2019, 06:04:46 pm »

Quote from: bob on January 21, 2019, 03:43:46 pm

You get the break when playing directly to a USB connected DAC??

I do, but in my case I am using MC 24.0.72 on macOS 10.13 using Core Audio to a USB-connected DAC.

The same behavior is seen with MC 24.0.74 on Windows 7 using WASAPI to an S/PDIF (coax) connected DAC.

From that this doesn't necessarily look platform-specific.

jac61gta · « **Reply #158 on:** January 21, 2019, 08:59:35 pm »

Hi Bob, yes thats correct. The behavior occurs in both scenarios i.e. using a direct connect DAC (usb) or DLNA. When the next song loads the current playing song drops sound (counter continues to count) for a few seconds, then resumes to end and the next song begins. The songs stream no problem otherwise.

bob · « **Reply #159 on:** January 22, 2019, 08:50:22 am »

What do you both have set in
tools->options->media network->advanced->Disable audio buffer to disk?
And what about the prebuffering under audio settings?

Shady Bimmer · « **Reply #160 on:** January 22, 2019, 10:35:05 am »

Quote from: bob on January 22, 2019, 08:50:22 am

What do you both have set in
tools->options->media network->advanced->Disable audio buffer to disk?
And what about the prebuffering under audio settings?

I don't see any 'Advanced' section under Tools -> Options -> Media Network on the macOS MC.
On the Windows 7 MC 'Disable audio Buffer to disk' is unchecked.

In both cases I have Tools -> Options -> Audio -> Prebuffering set to the default recommended '6 seconds'.
Tools -> Options -> Audio -> Memory playback is also left at 'No memory playback' on both instances.

Edit: I might also note that this problem does not always occur. For instance so far today I have not experienced this. Typically once it starts happening it will continue happening, however.

bob · « **Reply #161 on:** January 22, 2019, 11:23:24 am »

Quote from: Shady Bimmer on January 22, 2019, 10:35:05 am

I don't see any 'Advanced' section under Tools -> Options -> Media Network on the macOS MC.
On the Windows 7 MC 'Disable audio Buffer to disk' is unchecked.

In both cases I have Tools -> Options -> Audio -> Prebuffering set to the default recommended '6 seconds'.
Tools -> Options -> Audio -> Memory playback is also left at 'No memory playback' on both instances.

Edit: I might also note that this problem does not always occur. For instance so far today I have not experienced this. Typically once it starts happening it will continue happening, however.

I've only been able to get this to happen on windows with the disk buffer disabled.

On your MacOS MC, I assume that Media Network is disabled? The options won't show up unless it's enabled, they really should be moved out of that section...
You can enabled Media Network long enough to check those advanced options then turn it back off if you wish.

Shady Bimmer · « **Reply #162 on:** January 22, 2019, 01:33:08 pm »

Quote from: bob on January 22, 2019, 11:23:24 am

On your MacOS MC, I assume that Media Network is disabled? The options won't show up unless it's enabled, they really should be moved out of that section...

Yes it was disabled. I temporarily enabled Media Network on the Mac and confirmed the 'Disable audio buffer' is unchecked.

With the macOS instance I don't normally get the end of the song resuming after a break. The end of the playing song gets cut off, then there is a silence, then the next song starts. On the windows instance I have heard both scenarios occur.

bob · « **Reply #163 on:** January 22, 2019, 01:38:23 pm »

Quote from: Shady Bimmer on January 22, 2019, 01:33:08 pm

Yes it was disabled. I temporarily enabled Media Network on the Mac and confirmed the 'Disable audio buffer' is unchecked.

With the macOS instance I don't normally get the end of the song resuming after a break. The end of the playing song gets cut off, then there is a silence, then the next song starts. On the windows instance I have heard both scenarios occur.

The point at which the sound cuts off is actually the end of the file. It's off in regards to the track data since there is a crossfade applied at RP.

What's your internet connection speed and latency like?

Shady Bimmer · « **Reply #164 on:** January 22, 2019, 02:20:26 pm »

Quote from: bob on January 22, 2019, 01:38:23 pm

The point at which the sound cuts off is actually the end of the file. It's off in regards to the track data since there is a crossfade applied at RP.

What's your internet connection speed and latency like?

You mean the song is cut off in the middle of lyrics (singing), very abruptly? The songs definitely are getting cut short and not audibly fading into the next song.

Are you suggesting that RP is cutting off the end of songs? This only happens (that I've noticed) when using MC and not the RP web app. With MC when it starts to happen, it continues to happen but then the next day it will be resolved. The next time it happens I'll restart MC and see if that makes a difference.

I have 1Gbps symmetric internet via Verizon FiOS and wired 1Gbps ethernet to both the windows and macOS hosts.

jac61gta · « **Reply #165 on:** January 22, 2019, 04:05:00 pm »

Quote from: bob on January 22, 2019, 08:50:22 am

What do you both have set in
tools->options->media network->advanced->Disable audio buffer to disk?
And what about the prebuffering under audio settings?

Hi Bob, I can confirm Disable audio buffer to disk is unchecked also. This is not like a fade into the next song. Its abrupt and ends a few seconds before the song should, then the song resumes a few seconds more, then next song begins. My internet speed is 32GB down and 11 up with 11 ms latency. Issue occurs either wired or wireless.
Thanks

UncleMilt · « **Reply #166 on:** January 23, 2019, 11:40:33 am »

Quote from: jac61gta on January 22, 2019, 04:05:00 pm

Hi Bob, I can confirm Disable audio buffer to disk is unchecked also. This is not like a fade into the next song. Its abrupt and ends a few seconds before the song should, then the song resumes a few seconds more, then next song begins. My internet speed is 32GB down and 11 up with 11 ms latency. Issue occurs either wired or wireless.
Thanks

I too, have had this issue.
Issue is not sensitive to internet download speeds
Under Tools>Options>Audio modify "Track Change" setting from "cross-fade" to "gapless"

newsposter · « **Reply #167 on:** January 23, 2019, 02:16:46 pm »

The cut-out is happening on their Roku channel as well.

Additionally, I lost the audio stream, but not the photo stream, for about 15 minutes at 2 PM US Central time.

jac61gta · « **Reply #168 on:** January 23, 2019, 03:01:26 pm »

Quote from: UncleMilt on January 23, 2019, 11:40:33 am

I too, have had this issue.
Issue is not sensitive to internet download speeds
Under Tools>Options>Audio modify "Track Change" setting from "cross-fade" to "gapless"

Hi, thanks. I had read your earlier messages and made sure my settings mimicked those that you mentioned solved this issue for you, however it didn't work for me. I agree its not sensitive strictly to internet download/upload or latency as my speeds are quite good. Its does seem to have something to do with JRiver having to make the connection again to RP servers, though why this would cause the sound to drop I can't say.

Listening last night for 3-4 hours I can say that it happened to every song. The amount of silence before the sound was heard again (I say it this way because I believe the song continues as the song counter continues during the silence) was seemingly random, lasting from 1 sec to 10, but mostly ~3 seconds. As soon as the next song would pop up into the list, immediately the sound would drop, then current song would be heard again, fade out as normal and the song just added would fade in and play no problem until nearing the end, then rinse and repeat

I wonder if its a re-connection issue because as the just added song begins to play my network activity spikes from 0 mbs to ~17-24 mbs for a few seconds while the full song downloads (its playing while it downloads). Network activity is 0 when I hear silence and also zero when the song has fully downloaded and playing. Its spikes up for only a few seconds, depending on how long the song is (how big the file is).

Does anyone know the ip address of the RP FLAC server(s)? It would be interesting to see if a constant ping would keep the connection up and the issue goes away. Just thinking out loud.

Not sure any of this is helpful. I hope it is

Thanks

RoderickGI · « **Reply #169 on:** January 23, 2019, 04:38:56 pm »

Quote from: jac61gta on January 23, 2019, 03:01:26 pm

Does anyone know the ip address of the RP FLAC server(s)? It would be interesting to see if a constant ping would keep the connection up and the issue goes away. Just thinking out loud.

Don't do that. RP has no doubt a limited resource and can only handle so many connections. If you are constantly pinging it and keeping your connection alive, it is one less connection in the pool for other people to use. The idea is that you connect to RP, download what you need, disconnect and let the next listener get what they want. RP would have to start ignoring such attempts to be able to service all users, which may make things worse.

The problem could already be as simple as it takes longer than expected to get a new connection to RP to get the next track. That could be caused by any component in the path from you to RP, including your ISP. By all means, measure the connection time if you can, or ping or trace it for a while to see what the round trip is, but don't try to permanently hold a connection. Besides the above issues, it may get your IP Address blocked.

Shady Bimmer · « **Reply #170 on:** January 23, 2019, 06:03:28 pm »

Today I used another Windows 8 instance with MC 24.0.74 for a while and noticed some additional oddities:

First, I can confirm that the issues, when they occur, always happen just after the next song is added to the last. My settings here were the same as my Windows 7 instance (also S/PDIF -> coax -> DAC) except I had forgotten to switch track change from cross-fade (4s) to gapless.

Today the behaviors were not consistent song-to-song and were among:

Current song does not get interrupted, ends, then next song begins (no fade)
Current song does not get interrupted, next song begins, both songs play during transition (not fade) from one to other
Current song does not get interrupted, next song begins at same time then stops, first song ends, then next song begins (again) with no fade
Current song gets interrupted briefly, resumes, ends, then next song begins (no fade)
Current song gets interrupted briefly, resumes and next song begins at same time, next song stops, first song ends, then next song begins (again) with no fade

Today the "interruption" was typically brief, at most one second, but in the past it has been as long as several seconds.

I wasn't able to pay too close attention but was able to note the variations above. I didn't actually time it but in the cases where the two songs were playing together the second song did only play for a few seconds, possibly the same length as the fade. Note that when the two songs were playing together there was no actual fade - both were full level and it was very obvious.

jac61gta · « **Reply #171 on:** January 23, 2019, 09:20:23 pm »

Hi RoderickGI... agreed not a good idea to perform a constant ping. I should have been clearer when I said a constant ping. What I meant by constant was for a song or two to see if the issue corrects itself. Your points about being caused by any component in the path from RP to ISP are well made. A traceroute might be interesting... Thanks.

I noticed this evening that when the song resumes after silence it continues for exactly 10 seconds before the next song plays. Will need to monitor that a bit to see if it holds true.

Hi Shady... we share the same issue but for you it manifests itself in different ways. Mine is consistent from song to song. Weird. Its almost like there's a 'marker' in each song that indicates when the next song should be played/downloaded and the marker is out of wack.

Having no real knowledge of how this works, I"m just guessing and I'm sure thats not helpful so I'll stop

bob · « **Reply #172 on:** January 24, 2019, 12:18:01 pm »

Quote from: newsposter on January 23, 2019, 02:16:46 pm

The cut-out is happening on their Roku channel as well.

Additionally, I lost the audio stream, but not the photo stream, for about 15 minutes at 2 PM US Central time.

I was thinking that part of the issue might be the latency in downloading the next track and this seems to confirm it. It's possible they are short of bandwidth for the number of users.

bob · « **Reply #173 on:** January 24, 2019, 12:22:50 pm »

Let me re-iterate that the gap (if it occurs) occurs at the REAL file boundary which is NOT the same as the track boundary EXCEPT for the first song in a block (look at the Playback Range field in MC).
That's why it appears to gap before the end of a track.

When I look at the log, I see a gap in the time it takes for RP to start streaming the next track after the file request.

Shady Bimmer · « **Reply #174 on:** January 24, 2019, 02:48:12 pm »

According to MC the same track is still playing (IE: it hasn't moved to the next song) when this occurs. That would suggest that the same track is actually split between two "files" by RP, which would seem to fit.

IE: There is still time left when the song stops, and when it resumes it is still playing that track. Perhaps that is what is confusing. It seems odd that this only happens with MC and does not happen with the web app (at least not for me)

It sounds like you are saying this interruption near the end of the song is expected and normal. I may just have to go back to using the web app then

I only chimed in as someone else that was seeing the same issue reported in this thread. I may have missed it but I didn't see anywhere that it was indicated that the issue is known and is normal/expected.

bob · « **Reply #175 on:** January 24, 2019, 03:20:24 pm »

Quote from: Shady Bimmer on January 24, 2019, 02:48:12 pm

According to MC the same track is still playing (IE: it hasn't moved to the next song) when this occurs. That would suggest that the same track is actually split between two "files" by RP, which would seem to fit.

IE: There is still time left when the song stops, and when it resumes it is still playing that track. Perhaps that is what is confusing. It seems odd that this only happens with MC and does not happen with the web app (at least not for me)

It sounds like you are saying this interruption near the end of the song is expected and normal. I may just have to go back to using the web app then

I only chimed in as someone else that was seeing the same issue reported in this thread. I may have missed it but I didn't see anywhere that it was indicated that the issue is known and is normal/expected.

It's really NOT supposed to have an audible gap. If you run the log in MC you'll see that it does a request to pull the next "track" before the current one ends. The request is made at 2x MC's audio buffer ahead setting.
Assuming you have left that at the default of 6 seconds that should give you 12 seconds of time before the next "track" is requested and the current "track" ends for RP to get that track to you.
If your track transitions in MC are set to gapless you shouldn't hear a gap.

The web app accesses the "tracks" somewhat differently in a way that doesn't fit well into how MC views the playlist.

Shady Bimmer · « **Reply #176 on:** January 24, 2019, 04:43:32 pm »

That's what I would have thought is there should not be a gap, as others have indicated.

I have tried posting details on exactly the behavior I am seeing to perhaps diagnose and identify why this is happening.

You keep snapping back that this is known. I understand what you are saying about where the break occurs. . my behavior that I noted from yesterday may seem to conflict with that or perhaps not but that is why I thought it worth noting. It is also why I specifically noted that I hadn't switched to gapless and had track changes left to cross-fade. . .and still experienced an interruption at times.

Like I said - I'm willing to help but it seems like that isn't wanted so I'm moving on.

Edit: FWIW today with my macOS instance I pretty much had a short gap (less than 1 second) on every song. At some times as I noted earlier the song abruptly ended and the next song started without the previous song ending, but it was not common today (only about a half dozen songs over about four hours that I noticed)

swiv3d · « **Reply #177 on:** January 24, 2019, 04:54:16 pm »

When I play RP it is pretty seamless.

swiv3d · « **Reply #178 on:** January 24, 2019, 05:31:48 pm »

Under audio you must have

memory playback: No memory playback ;

Do not play silence leading or trailing -deselected

this gives me flawless playback of RP.

Shady Bimmer · « **Reply #179 on:** January 24, 2019, 05:49:01 pm »

Quote from: swiv3d on January 24, 2019, 05:31:48 pm

Under audio you must have medory playback: No memory playback ; Do not play silince leading or trailing -deseected this gives me flawless playback of RP.

Already there and confirmed that (see earlier posts). Memory playback is disabled by deafault and has been confirmed to be the case on all three of my instances.

I don't dispute that many (most) have no issues at all. If this was something consistent and easily repeatable (IE: everyone saw this) it would be something that could be more likely be corrected or otherwise mitigated.

I also don't dispute where this is happening. . . I determined that was likely the case pretty much when I first experienced this. The challenge now is to identify why.

I see slightly different behavior between Windows 7 instances and macOS instances. Possibly noteworthy is that in all cases 64-bit builds are in use.

I've spent enough time on this but the responses don't seem to be appreciative so I'm not spending more time to try to identify details that might help with resolution.

Shady Bimmer · « **Reply #180 on:** January 24, 2019, 06:51:36 pm »

The settings you recommend had already been confirmed. .

bob · « **Reply #181 on:** January 24, 2019, 07:31:55 pm »

I think you are misinterpreting the tone of my messages at any rate. I’m not trying to discredit what you are seeing. Just trying to figure out how it can happen since with the settings you are using I cannot reproduce this on an internet connection way slower than yours and with machines slower than yours as well
I can only make the gap happen on raspberry Pi Linux running MC, orders of magnitude slower than the average PC and even then the gaps don’t always happen using the latest MC 24.0.74.
You aren’t using a vpn are you?

Shady Bimmer · « **Reply #182 on:** January 24, 2019, 08:06:59 pm »

No VPN here.

Shady Bimmer · « **Reply #183 on:** January 24, 2019, 08:27:02 pm »

FWIW - I don't know if this is necessarily network related. I'd have been open to suggestions to help confirm or rule that out but at this point I've lost interest. I don't know how MC handles cross-fade with track changes when streaming RP, but the behaviors I noted above are puzzling to me at least, based on what has been posted in this thread. The gapless track-changes on macOS also have been puzzling, where the current song ends abruptly rather than being resumed after the break.

jac61gta · « **Reply #184 on:** January 24, 2019, 09:27:51 pm »

Quote from: swiv3d on January 24, 2019, 05:31:48 pm

Under audio you must have

memory playback: No memory playback ;

Do not play silence leading or trailing -deselected

this gives me flawless playback of RP.

Hi, thanks but unfortunately these setting do not help in my setup. Curiously, I don't think these settings are necessarily relevant since I'm experiencing this issue both directly connected to DAC via USB and via DLNA, which doesn't provide for these settings as JRiver audio engine isn't being used when DLNA is employed. Hmmm

I do note that this issue is consistent across a few different machines here running Win10 and one machine on Server2012. The behavior is consistent across the board i.e. sound cuts out. I tried switching to another piece of software using the RP plugin and no issue, however this necessitates using different hardware. Not sure what this means, just adding data points.

I'm enjoying 24.74, sounds great, playing around with panel and kinda getting used to the drop out issue lol... though would be nice to transition properly

jac61gta · « **Reply #185 on:** January 24, 2019, 10:08:23 pm »

Hi Bob, I was reading your notes over a few times and its interesting that when you say '...the log in MC you'll see that it does a request to pull the next "track" before the current one ends.' this is exactly when the sound drops for me, consistently. I'm not reading the log, but in MC when the next song drops into the list as next to play thats when the sound drops. Its immediate and repeatable every time.

Shady Bimmer · « **Reply #186 on:** January 24, 2019, 10:16:03 pm »

I concur with the above. When I've been watching any interruption happens immediately right after MC adds the next song to the playlist, as if there is no buffering. The cases on my windows instance where two songs were playing together (with or without interruption) also happened at this time. MC shows the same track still playing (after the break) for the last few seconds before moving to the next song.

On macOS, it is more difficult to follow, but when I've been actively watching the break happens at the same time, but the resumption with the next song is shown as a new track altogether.

jac61gta · « **Reply #187 on:** January 24, 2019, 10:43:11 pm »

Hi folks... another data point that is consistent for me. Looking at Task Manager, when the song is playing I show:

CPU - ~24%
Memory - 56%
Disk - 0-4%
Network - 0%

But as soon as the music gaps i.e. the next song is added to playlist, the 'Disk' red lines to 99% with System and Media Center 24 (32bit) at the top. All others remain the same. When 'Disk' drops back down the music resumes (or is it the other way around?). Then 'Network' increases as the next song downloads (plays at the same time), as one would expect.

JimH · « **Reply #188 on:** January 25, 2019, 07:54:33 am »

I've removed some off topic grumbling.

bob · « **Reply #189 on:** January 25, 2019, 08:20:12 am »

What it's doing on that transition is buffering the next track to disk as fast as your internet connection will serve it which is why the cpu usage spikes (this is only a single thread that has less priority than the playback thread so it still shouldn't cause a gap). That's with the setting for disk buffering (under Media Network->Advanced) DISABLED which is the default.

When that option is checked, the disk buffering is skipped and the next file is seeked (re-opened) by the internet reader code.

It might give a clue if those that can reproduce this can see if there is a difference between the two methods of retrieval.

Another thing I though of last night, we all have SSD's in our computers here, perhaps some of you have hard drive based machines and the disk is sleeping?

On the Mac at any rate I know that's a setting in under System Preferences->Energy Saver->Put disk to sleep whenever possible. I'm sure there is an equivalent in windows.

Also, last night and this morning I played RP on a 7 year old Mac Book Pro (the original HD was changed to SSD) and it played flawlessly on my pretty modest wireless network.

I'd really like to figure this out so any more data points would be useful.

jac61gta · « **Reply #190 on:** January 25, 2019, 05:01:35 pm »

Hi Bob, not sure if a typo or not but when you say cpu usage spikes, did you mean disk spikes? My cpu never goes above ~35% during this issue.

Disk buffering on/off makes no difference here, it gaps either checked or not.

You may be on to something regarding SSD and normal hard drives. I have all hdd here. No settings I can see in WIN10 where the disk sleeps, plus it doesn't look like my hdd is sleeping. I did try enabling Disk Cache Writing in WIN10 setting and this made no difference. Also played with power plan management to no avail.

The attached screen shot shows the disk red lining at 100%. This happens immediately/instantaneous when the next song shows up as next to play in MC (hasn't downloaded yet), then gap ends when disk performance returns to normal, sound returns, current song continues to play for 10 seconds (ends/fades out normally) then segues into the next song that was just added, which then begins to play and downloads.

JimH · « **Reply #191 on:** January 25, 2019, 05:34:37 pm »

Antivirus?

jac61gta · « **Reply #192 on:** January 25, 2019, 05:37:57 pm »

Bob,
Not sure if I'm misunderstanding or not, or how important it is but, when you say:

What it's doing on that transition is buffering the next track to disk as fast as your internet connection will serve it which is why the cpu usage spikes (this is only a single thread that has less priority than the playback thread so it still shouldn't cause a gap).

When I get the gap, the transition has not happened, by this I mean the next song, while added to, and seen in, MC as next to play, has not downloaded yet. My network activity is zero. When that song starts to play then and only then does my network activity show ~25 Mbps receive for ~5-6 seconds (song plays while it downloads). So if I understand correctly its not really buffering when the gap occurs?

Hi Jim, hope all is well... yes I certainly thought of that. I use McAffee. First thing I did was shut it down, gap persisted.

Its funny, I'm almost getting used to the gap lol. I'm more curious than anything

JimH · « **Reply #193 on:** January 25, 2019, 05:42:47 pm »

Uninstall the antivirus to test.

bob · « **Reply #194 on:** January 25, 2019, 08:14:31 pm »

Quote from: jac61gta on January 25, 2019, 05:37:57 pm

Bob,
Not sure if I'm misunderstanding or not, or how important it is but, when you say:

What it's doing on that transition is buffering the next track to disk as fast as your internet connection will serve it which is why the cpu usage spikes (this is only a single thread that has less priority than the playback thread so it still shouldn't cause a gap).

When I get the gap, the transition has not happened, by this I mean the next song, while added to, and seen in, MC as next to play, has not downloaded yet. My network activity is zero. When that song starts to play then and only then does my network activity show ~25 Mbps receive for ~5-6 seconds (song plays while it downloads). So if I understand correctly its not really buffering when the gap occurs?

Hi Jim, hope all is well... yes I certainly thought of that. I use McAffee. First thing I did was shut it down, gap persisted.

Its funny, I'm almost getting used to the gap lol. I'm more curious than anything

Yes, the spike in disk/cpu is the thread writing the next track to disk. You shouldn't see that at ALL if that disable buffer to disk thing is checked. If you do, something else is going on.
On a windows machine, even microsoft's security essentials will scan the temp file being written to disk.

RoderickGI · « **Reply #195 on:** January 25, 2019, 10:30:14 pm »

I hear the pause occasionally. Not on every track, but enough to be able to watch what happens and when, if there is a glitch.

I just did a listening test again and watched the disk and network activity. My temp directory is on my OS SSD. I had a backup of one hard drive running to another hard drive at the time, so one hard drive was maxed out. CPU stayed below 20% with no significant spikes. All this other activity seemed to have no effect on the occurrence of the glitch. This was on my Workstation so not the PC in my signature. A 2010 era PC, but still a powerful PC.

I have "Disable audio buffer to disk" unchecked. Audio prebuffering at 6 seconds. Internet speed 55 Mbps down, 5 Mbps up, latency 9 to 16ms. Ping to www.radioparadise.com ~190ms. No memory playback. Gapless. Playing to Realtek High Definition Audio (WASAPI) Exclusive, 100ms buffering.

As far as I could tell, the glitch happens not when the download starts, and not when the disk activity happens, but just before that when there is send activity on the network. I would guess that is when MC sends a request for the next track to Radio Paradise. I was almost able to predict if a glitch would happen before the download or disk activity started, based on whether the quick network send completed without a glitch occurring. The process is quick though, so my predictions weren't 100% accurate. Also, sometimes the send happened at the same time as some other network activity. Possibly even the track download, which seemed to start immediately when the send happened, rather than shortly afterwards.

So maybe look at the code that is requesting the next track Bob.

PS: I did start to look at what was going on using Process Explorer, but by the time I worked out how to detect and capture a glitch, Process Explorer decided to crash. I might have another look at it later. Maybe someone else might try using Process Explorer to zero in on the cause.

jac61gta · « **Reply #196 on:** January 26, 2019, 04:04:30 pm »

Hi RoderickGI... thanks for adding your observations. The more data points the better.

Jim / Bob - I removed / deleted my McAfee virus software, rebooted, and still no joy. When I did so I noticed Windows Defender turned back on (makes sense) so I completely shut it down. Not sure how, or if even possible, to delete it though.

bob · « **Reply #197 on:** January 27, 2019, 07:11:52 pm »

I added some read debugging to the log so that we can tell how long after the seek the read begins.

I hope that helps track down the issue...

UncleMilt · « **Reply #198 on:** January 29, 2019, 04:14:12 pm »

Quote from: RoderickGI on January 25, 2019, 10:30:14 pm

I hear the pause occasionally. Not on every track, but enough to be able to watch what happens and when, if there is a glitch.

I just did a listening test again and watched the disk and network activity. My temp directory is on my OS SSD. I had a backup of one hard drive running to another hard drive at the time, so one hard drive was maxed out. CPU stayed below 20% with no significant spikes. All this other activity seemed to have no effect on the occurrence of the glitch. This was on my Workstation so not the PC in my signature. A 2010 era PC, but still a powerful PC.

I have "Disable audio buffer to disk" unchecked. Audio prebuffering at 6 seconds. Internet speed 55 Mbps down, 5 Mbps up, latency 9 to 16ms. Ping to www.radioparadise.com ~190ms. No memory playback. Gapless. Playing to Realtek High Definition Audio (WASAPI) Exclusive, 100ms buffering.

As far as I could tell, the glitch happens not when the download starts, and not when the disk activity happens, but just before that when there is send activity on the network. I would guess that is when MC sends a request for the next track to Radio Paradise. I was almost able to predict if a glitch would happen before the download or disk activity started, based on whether the quick network send completed without a glitch occurring. The process is quick though, so my predictions weren't 100% accurate. Also, sometimes the send happened at the same time as some other network activity. Possibly even the track download, which seemed to start immediately when the send happened, rather than shortly afterwards.

So maybe look at the code that is requesting the next track Bob.

PS: I did start to look at what was going on using Process Explorer, but by the time I worked out how to detect and capture a glitch, Process Explorer decided to crash. I might have another look at it later. Maybe someone else might try using Process Explorer to zero in on the cause.

My Radio Paradise "playback pause or gap" has disappeared altogether? I have removed gapless playback transition options and still no symptoms? Song transition shows ~7 second buffering but zero silence or playback interruption.

What has changed? Most grateful regardless!

bob · « **Reply #199 on:** January 30, 2019, 10:51:50 am »

Quote from: UncleMilt on January 29, 2019, 04:14:12 pm

My Radio Paradise "playback pause or gap" has disappeared altogether? I have removed gapless playback transition options and still no symptoms? Song transition shows ~7 second buffering but zero silence or playback interruption. What has changed? Most grateful regardless!

It really could be on their end.
If you have gapless turned off, what is the track transition set to?

INTERACT FORUM

Author Topic: NEW: Radio Paradise (Read 58017 times)

swiv3d

swiv3d