[Hersan]

Hi

A few small questions regarding Frodo. The options for video scaling metod in Frodo (nightlies) have been reduced to only show billinear or nearest neibor (In Eden I use Lanczos3 optimized). Perhaps this "downgrade" in scaling metod is small, but I noticed that the video quality in Frodo is significantly worse than Eden on my machine. I havn't been able to test Frodo until now, the hardware acceleration have not been working (giving me green screen, stutter etc), but now it seem work (some what).

Apart for the above video quality issue, I see that cpu activity is higher in Frodo than in Eden, 23 % on average compared to 9 %. (hardware acceleration is activated on both).

Slightly connected to the above: (in Eden) starting a video (720p mkv) results in severe judder, stoping it and starting it again solves the problem (the issue appeared just a week ago and must have something to to with me trying pre-frodo because Eden worked just fine before).

Further more (on my macbook pro retina) the "home-screen" of Frodo (not Eden) looks very "smeared or water-coloured". Homescreen (background) looks fine when connecting to my TV (1080p) and blanking the laptop. Must have something to do with the extreme resulotion and how the mac handles it. No biggie.

I have scanned through a large part of the forum but I can't seem to find anyone who have raised these questions before (at least not so too the point).


Perhaps I should have started with a big thank you for a superb program! I use it alot I can tell you!

Setup:
Macbook pro retina (OSX 10.8 ML) ->hdmi to reciver-> hdmi to TV

BR
H

[davilla]

Lanczos3 was removed from osx because most users would select it even though their boxes did not have the CPU ponies to do it. Video playback would jutter/stutter/stall and this would cause a large amount of frustration. In addition VDA does not use these options as it is handled in a different way driven by what comes out of the hardware decoder.

I can't comment on cpu activity as just stating the percent without the context of what you are doing means nothing

Also since you don't mention which nightly version you are using, VDA issues with decoding under ML was fixed a few days ago.

I cannot stress this enough, we cannot help anyone that does not at least indicate which version (exact version) they are using. This info is and has always been the 1st thing in xbmc.log. It's there for a reason... so we know exactly which version is running. As far as I know, you are using a version that is several months old.

[Hersan]

Hi Davilla and thanks for your reply.

Regarding your last comment, your are absolutly right. No log, no help for obvious reasons. However I just posted as a quick question and I'll type back after more testing (and with logs) regarding the cpu usage.

But so I understand it correctly. VDA (hardware acc) dosn't use any of the (in Eden) stated scaling methods, thus using VDA in Eden would result in the same picture quality as Frodo (with VDA) (all else the same)? This was my quess, the sad part is that VQ is worse on my machine with frodo compared to Eden (this is why I started to go though a settings etc to see if I had checked something I shouldn't). No judder or anything, with worse I mean less picture detail, in a way less sharp. Odd right?

Mostly I play 720p (mkv) videos and use xbmc to upscale it to 1080p before sending (through my reciver) to my TV (all with hdmi). This is why I mumbeled about scaling methods...


From your perspective (since I'm all about pushing picture quality to the limit), would not using VDA and instead switch to software (with a more advance scaling method) result in better VQ? Or is it marginal?

About the nightly, again sorry for not pointing it out, the ones built on the 14th of august, 64bits, thus including the VDA fix.

Best
H