Agree with Ian_T. Nothing wrong with patches that push the boundaries, if anything this might be a good starting point to work backwards from. Even though it didn't work for me it is pretty stunning and hopefully someone can reverse it to find a more stable setting that still maintains that picture quality.
To push boundaries, risk is inevitable. No risk equals no gain. A hack that works 99.9 % of the time will still be consider as not production ready for some professional. Today I use Driftwoods 176M GOP1 for a paying job. Was I worried about the stability of the patch on the shoot? Of course I did. I just kept reminding myself to review the important shots on location with a laptop to calm my nerves. For safety, I have another unhack GH2 as a standby cam. At the end of the day, the hacked GH2 won big time with the Driftwood 176M GOP1. The unhack GH2 lost. I took the risk and I was rewarded with beautiful images and praises from the client. So I told my client, "See I told you my Lumix GH2 is a special edition, it's not the same GH2 your wife uses."
There is a big difference between 90% reliable and 30% reliable. This patch falls in the latter category. @Driftwood 176M is very safe. Shot many hours of footage with it under all circumstances, and also survives my motion deathchart test.
If it were not for the guys here looking for the best and and determing our limitations, this would all be moot. IMO, this is the info Vitaliy and Chris want to hear.
sohurs, is of course right when importance and $$ is the main factor. If taking chances, you have no choice but to check it in the field.
We just did a few low light tests between this and driftwoods 175. Those saying it's a crash maker are right, but it's also at another level in the shadows when kept under 400 iso (we were getting crashes above that). My impression was that the shadows looked like they'd been shot with a faster stock, just a tighter, cleaner grain.
We watched "Il Pranzo di Ferragosto" last night, and I kept thinking, while watching all the shadow play in that dreary apartment that these are the shots we never show when showing off the GH2, drab underlit shots rolling from room to room. But in this film the DP created the mood with these elegant shadows. Henry's patch is getting close to letting us really embrace that kind of methodology with the GH2.
That all said, if we could find a happy medium between this and driftwoods reliability, that would be the ideal for me. In the meantime, when there's some light and darkness to be played with (and enough of a baseline that we can keep below iso400) I'll be experimenting with this new patch. Thank you for pushing things, Henry.
Hey Henry, a valiant effort and that patch that will have its uses. Hopefully we can set some clear parameters with this patch, situations it's suited for and NOT suited for.
Analogy: a 800mm f5.6 lens can be awesome, provided one knows where and when to use it.
Man, I just wish there was a way to have these high bitrate high IQ settings and set limits that could maintain buffers and hold cadence. This is a 180M GOP3 AQ4 setting. Has someone tried this set to AQ3 and checked if it's more stable?
We just tried at AQ3, and while the setting still looks really good, now it's about level with the amazing 176 driftwood, and still creating much larger files, so I'm not sure it's worth it at AQ3. I think this is a have it in your back pocket but don't use for everyday specialty setting (at AQ4).
@Ralph_B - if you're ever going to do anymore a/b HDMI vs AVCHD tests, please consider trying Henry's patch. You'll need to shoot at a lower ISO (we've been good under 400), but I think the grain will be closer to the HDMI than any comers to this point.
Edit - try at AQ4 as he has it. Not much benefit at AC3 that I can tell.
@HenryO - What are you doing to transcode these files into an etitable format? We're finding clipwrap and FCP log and transfer can't seem to deal with these files and sputter out. Any ideas?
@HenryO Cadence bigtime if you manage to get it recording for more than 5 or 6 seconds. Mate, you seriously cant reccommend this to people - theyre gonna have problems on the edit when they get their rushes back off this setting. You may get 2 to 5 seconds of good takes only and the file sizes are no bigger than whats already out there. The chart below was a 20 sec max shoot I managed to get out of it shooting interior. PS it takes 2 seconds for this to die on pappas death chart.
@Driftwood I completely agree. See my earlier comment. This is a fail patch. I don't care if it excels a few seconds. This patch is 99% unreliable, even in controlled settings. You can go out and shoot, and might be lucky, but that one day that there is more detail in the background... you get a clip like this. Not for me. I wouldn't mind if this was posted in extreme settings though, but to announce this as the next big thing in GH2 hacking land... not smart.