FX1 HDV artifacts that bad?
 

Looking at these images http://www.dvxuser.com/V6/showthread.php?t=46255&page=2

I wonder how the Fx1 is capable in hard compression situations, It is holding me off now for using the FX1/Z1 for sportshooting with lots of motion, is the realworld that bad? Those HDV artifacts are so visible.

I had the same questions after I bought my FX1 I don't believe there is a real problem... see this thread.

http://www.dvinfo.net/conf/showthrea...hris+barcellos

We regularly shoot college basketball floor level.

Haven't seen any yet.

I can't vouch for sports footage, but so far I've never seen an objectionable compression artifact on my Z1, even when I've tried to create one. But then I don't go shooting exploding ships at night, which is always going to look bad with MPEG-2 compression.

In fact, if the final format for your footage will be MPEG-2 (e.g. DVD or broadcast HD), then it probably doesn't matter what you shoot on: if it would have horrid artifacts on the FX1, odds are it would have horrid artifacts if you shot IMAX-3D and then compressed it to MPEG-2 for delivery. I've seen broadcast HD footage shot lovingly on $100k HDCAMs which looked far worse artifact-wise by the time it hit the airwaves than anything from my Z1, and entire movies spoilt by MPEG-2 compression artifacts on broadcast digital SDTV: one movie set on an old sailing ship where the entire picture broke up every time a wave crashed over it, for example.

Alot of times this is because of hardware encoders passing the footage through blind, where a software encoder can take it's time to get around blocking by allocating bits better.

Luckily some blocks can only be apparent if you step frame footage.

Douglas Spotted Eagle has some 60i clips of motocrossing on the Vasst site. You should view them.

That said, my $0.02 is that motion is handled well, artifacting is under control, and that judgments from viewing interlaced clips on pc monitors may lead to erroneous conclusions. Try them on a decent HDTV monitor instead.

A huge factor in the resulting quality of MPEG compression is what's called "Motion Search Precision" (at least that's what TMPGEnc calls it). If I understand correctly, it's basically about how wide the search is for similar blocks in adjacent frames, for the interframe compression (where similar blocks are found, those areas of the image are essentially copied, by reference to a block in another frame, rather than separately encoded). Real time compression (such as on-camera while shooting) is limited in the amount of time that can be dedicated to this search (the codec has to keep up to the speed of acquisition). Where similar blocks in adjacent frames can't be found, all of the information for that part of the image has to be encoded, essentially the same as if it were an I-Frame. When compressing to create an MPEG stream for a DVD, there is the luxury of taking more time to do motion search thoroughly (for high quality output, always make sure the setting is at max for Motion Search Precision - and then go fishing while your computer churns).