Can't see why you need to remove art anyways. I'd try just highlighting the tracks, copy it from the web and then pasting the art in (or loading it from a file works too) ... even if I'm changing the art on a track I'm currently listening too, it just won't write to the file until its done. Try it and see. Thats why I think it has nothing to do with memory or latency at all. Probably shouldn't change those settings anyways as they are probably set-up optimally for normal playback.
It's always been a "completest" thing with me
When I change album art - the old has to go before the new goes in. So I remove everything to be sure and then lay in the new stuff. And yes - I am aware that MC will not write to the file that is currently playing - but that's not what I am doing. I am playing a track either from the network location OR my local drive - that has nothing to do with the art I a going to change. While this track is playing - if I go to a completely unrelated album and change it's art - it's glitch city.
I ran some logging against this process (dropouts/glitches) and MC reported NOTHING with respect to any interferance with the actual audio stream throughout the entire logged event (changing artwork on a selected album in the library). I did notice some real weird stuff going on when the art was about to be added to the file:
0012527: 4544: General: CFileInfo::ImportCoverArt: filename=D:\Pictures\To Be Filed\XTC - The Big Express (1984).jpg
0012527: 4544: General: CFileInfo::ImportCoverArt: Finish (0 ms)
0012527: 4544: General: CFileInfo::Close: Start
0012558: 4296: Reader: CWinINetReader::Open: Start
0012558: 4296: Reader: CWinINetReader::Open: Opening
http://www.yadb.com/cgi-bin/CoverArtExists.cgi?High=3386185246&Low=4308396120012558: 3796: Reader: CWinINetReader::Thread: Start
0012558: 3796: Reader: CWinINetReader::Connect: Start
0012558: 3796: Reader: CWinINetReader::Connect: Finish (0 ms)
0012558: 3796: Reader: CWinINetReader::DownloadFromHTTPURL: Start
0012745: 3796: Reader: CWinINetReader::DownloadFromHTTPURL: Success
0012745: 3796: Reader: CWinINetReader::DownloadFromHTTPURL: Finish (187 ms)
0012745: 3796: Reader: CWinINetReader::Thread: Finish (187 ms)
0012745: 4296: Reader: CWinINetReader::Open: Finish (187 ms)
0012745: 4296: Reader: CWinINetReader::Read: InternetReadFile returned true with zero bytes read: EOF
0012745: 4296: Reader: CWinINetReader::Close: Start
0012745: 4296: Reader: CWinINetReader::Close: Finish (0 ms)
0012745: 4296: Reader: CWinINetReader::Open: Start
0012745: 4296: Reader: CWinINetReader::Open: Opening
http://www.yadb.com/cgi-bin/CoverArtSubmitFile.cgi?High=3386185246&Low=4308396120012745: 4284: Reader: CWinINetReader::Thread: Start
0012745: 4284: Reader: CWinINetReader::Connect: Start
0012745: 4284: Reader: CWinINetReader::Connect: Finish (0 ms)
0012745: 4284: Reader: CWinINetReader::DownloadFromHTTPURL: Start
0013198: 4544: General: CFileInfo::Close: Finish (671 m
MC preps the target file for write access and
then ducks out to the internet (YADB) for some bizarre reason before finally writing my image to the file. What exactly MC needs to be going to YABB for during this process is concerning to me. I am writing my own images to my own files and I do not see any reason or point to go to YADB to "check" for something.
This bizarre process uses my network card too - making for even more possibility of contention. Not to mention a privacy or rights issue as well. Is YaDB taking a copy of my artwork without telling me? Looks like it might be with a file named CoverArtSubmitFile.cgi?
I would like more information on exactly what the Add From File process is doing - becuase this does look a little weird. I would expect the process to open my file, write my image and get lost. Not venture out to the internet everytime I want to add some art for MY music files.
That aside - during this log period - MC was glitching constantly as the image file was being written to the 10 tracks for this album - but the log records no oddball behavior for the dropouts. Which tells me the MC is not even aware that it's glitching. Making this look more and more like a hardware (IRQ) issue as before.
Will keep on digging - but I do not like the look of that YaDB stuff in this log.
VP