paglez

In cinematic motion I refer how motion is captured, not in picture quality in itself: http://www.personal-view.com/talks/discussion/2848/gh2-in-24p-noticeable-strobe-effect-with-any-motion-in-frame/p4?Sort=date

@mo7Ies 24p Cluster v4 is good and in my own motion water and seas test stays close to target Q 16 achieving better results than flowmo. Lee's test above is carefully massaged to show better results. He would make a wonderful politician but I'm not sure I would trust his Lab results from now on.

Give Cluster v5 /5.1 a try as I want your feedback - lets get this setting nailed.

@driftwood I was using the Panny 45-200. Bright sunlight with no ND so shooting 100-200 shutter with 7.1-11 fstop mostly at the far end of the lense. Tripod mounted but pretty fast pans as trying to catch fast action. I do remember a lock up when I was using auto focus continuous. There was a wire fence in the BG and it locked on that because I think the auto focus probably got confused. I had a couple of lock ups more than when I used Flowmotion the week before. Looks bloody good but lock ups in 720p Pal mean I need to look at a lower quality but more reliable patch perhaps

@rsquires Ok, thanks for testing. Don't worry Ill adjust a bit more balance for you. Keep the results coming chaps. We'll get reliability in the end. I will be doing a Crossfire Cluster of the best of all settings once we establish the all best performance.

@rsquires I expect the difference is that I was shooting in lower light (Sun below the treeline), and doing slow pans. The bitrate was pretty low as I recall (fiftyish). I had the same problem as you with Sedna. Super images but lockups in fast pans especially over water and bright light.

I was also in Shutter priority and an AE/AF lock most of the time at 1/125.

@LPowell...you're both bloody geniuses Lee!

@rajamalik One of the reasons I keep them off is because they are very unpredictable about when they do and do not kick in. Several stills photographer reviewers on different sites commented on how they were unable to reliably get one of them to kick in - it sounded like that was the case even when they stayed in controlled environments, but I may be inferring too much.

However, if you go look at the various reviews (I think CameraLabs and DPReview may have been two that had examples) you can see what the stills look like with and without.

As for me, I leave them off so that I can have consistent results.

@Driftwood I do not think LPowell massaged his results. The scene tested is the one he normally uses as a torture test and he appears to have been making an effort to optimize the setting to handle it for some time. It would be natural for Flowmotion to be able to handle it well given comparison to other settings within a similar GOP and bitrate settings and at the same time there can be a lot of variation in water scenes.

This does not mean that the results will be consistent in all other types of scenes.

Your settings are great, and I use them the overwhelming majority of the time. But just because the two of you are getting different results in your testing does not mean that either of you are doing any massaging and LPowell has not done anything in the past to draw suspicion - quite to the contrary. :)

Honestly, the best comparisons (for all settings authors) should be done independently. Trying to be fully objective about comparisons of your own settings can be like trying to be objective about someone rating artwork by your kids. :)

@rajamalik http://personal-view.com/talks/discussion/1110/idynamic-range-possible-to-hack/p1

Globular v.5 does not seem to handle fine detail well - but I do not know how much mac-based software is involved. But I have the same glitch in VLC, Realplayer & Quicktime: Image one shows you the filming situation (somebody painting a canvas) , Image two shows the glitch some kind of "draining watercolour" effect (this was NOT the artist), when Canvas got in focus and shows fine detail. 24H, Nostalgic, all-2 except colour +2 / 14-140 Iso Auto, Shutter priority

@driftwood I'm not "spamming you", I'm replying to direct comparisons between Cluster and Flow Motion v2 that @mo7ies posted earlier in this thread, as well as your bogus claim I "carefully massaged" my test data. As a rebuttal, I've uploaded the actual, unedited footage to enable anyone who wishes to download and evaluate themself. I'm also calling bullshit on your simplistic "my numbers are better" sales pitch. Instead of these petty attempts at one-upmanship, how about posting some real footage and let that speak for itself?

@LPowell People can look & test for themselves. Stop repeat posting your tests recorded at different times. Like I said I merely mentioned the differing light between the two) My tests were conducted at the same time on two different cameras carefully pointing at the same subject matter, conditions etc...

There's no sales pitch either I've demonstrated detailed fact using Chris Brandin's Streamparser.

I really don't want to continue this sillyness.

@driftwood Movist might help BUT: I have the same glitch in FCP7 ,Clipwrap, Premiere 6. Maybe someone else could also check fine detail.