Ok you spotted the difference/ However, the Mysteron clip was shot at 800 ISO f5.6 shutter 100 - The RAW at 400 ISO f4 1/40 shutter.

This thread is about to close... new thread going up as we have over 1000 comments again...

The new thread link 4 is at:-

Series 4 Link

PLEASE DO NOT POST ANY MORE COMMENTS HERE. VK IS ABOUT TO CLOSE THIS THREAD.

Noise isn't always bad anyway - by which I mean, if you have a subject where you get a lot of banding (smooth gradation of light falling on a whitish surface, for example) then a noisier setting which you then noise-reduce in post will pretty much eliminate the banding, whereas if you have a very clean shot you will be stuck with banding.

I was experimenting with a Nikon 1.4 at a lower ISO (giving very clean video) and the banding on the tall candle was horrendous as the candle in the cage (on the left side) flickered. I went back and shot it at a much higher ISO and denoised in post, and the attached resulting image has almost no banding. You can still see some, but believe me, the low-ISO version was really, really bad whereas this noise-reduced version is pretty much fine.

So noise can be your friend!

@ishvar "There is however quite a lot of mosquito grain in Driftwoods patches fuzzing around @ higher iso's "

I guess you refer to normal sensor noise, not mosquito noise? While sensor noise is unavoidable and will get more visible the higher you set the bitrate - cause noise is details and low bitrates somewhat "denoise" by simply trowing this detail away - mosquito noise is the result of heavy compression. This is a big difference, sensor noise can be cleaned very well with Neat Video or MB Denoiser, while mosquito noise is a bitch in this respect.

In short: If a setting gives you alot visible sensor noise thats a feature, if you see mosquito noise, its a bug.

(And yes, I know there is things going on with matrixes and deartefacting that do play a big role in how noise shows up in the compressed stream - but we are quite far away from visible mosquito noise IMHO.)

@ishvar @Stray In my experience, it is the higher bitrates (not the GOP length) that taxes system playback. For instance, SpanMyBU v1 plays back on a PS3 much more easily than Orion does.

@ishvar Panasonic is not involved, it was an April Fools joke. Although I do a lot of testing with the GOP1 settings, for any work/project that I do I always use a GOP3 setting. Personally, although I do have a pretty fast PC, I've never experienced any additional hassle or overhead working with GOP1 footage over longer GOP footage. Rendering and grading should actually be a faster process as the applications can build frames quicker (the reality is no real noticeable difference on the whole), grading is also a lot more flexible with these lower GOP settings.

ISO Bug after sleep mode is VERY bad news, I did not recognize that before. For larger projects we run a external V-Mount anyways and disable sleep mode completly, but for the small stuff, jumping around without the heavy rig this is a real drawback. It means you either have to do 5 clicks each time the cam wakes up (yes, dangerous to forget) or disable sleep mode and empty your batteries in a quarter of the time.

This should be top priority to fix for Panasonic! I don´t understand at all why they did not adress this in the firmware update.

However, back on topic: Sedna is really amazing, I did not try Mysteron until now, but Sedna is already soo freaking good, even for the big screen. Thanks alot!

@Kihlian Haven't read that thread yet. Will check it out soon but am busy with the music video stuff right now. :)

http://www.personal-view.com/talks/discussion/2788/joanna-st.-claires-new-single-prayer-of-light-music-video-shot-in-part-with-gh2-sedna-aq1-c#Item_6

(The one that mixes Sedna AQ1 C with HV30 footage and heavy grading.)

@GH2_fan Best of luck with you finding one that works. I wish I could more help. :)

@GH2_fan just to give case and point. I had a 45mb/s sandisk 16gb card that had error after error even with some of the more gentle patches. I exchanged it for a 95mb/s card which just arrived yesterday (after about 3 1/2 months waiting). The 95mb/s card is working without flaw so far including my test last night of moonlit cherry blossoms at 6400ISO using mysteron on 1080p.

@GH2_fan The problem with the 45MB/s SanDisk cards is that there are several versions. Some of the versions are discussed here.

http://www.personal-view.com/talks/discussion/90/sandisk-extreme-sdhc-cards#Item_301

Some of the SanDisk 45MB/s perform well, other models less so. If you get the 95MB/s SanDisk, there is currently only one version and it is the best-performing card that has been tested so far. So if you want peace of mind, get the SanDisk 95MB/s but if you find a great deal on the 45MB/s version, make sure it is the right one (check in that link I included).

@GH2_Fan @jvalal The SanDisk 64GB 95MB/s will run all the new patches and has the highest probability of spanning of any card tested so far. I have also tested the Delkin 16Gb 95MB/s and it has worked with every new patch in 24H (I have not tried Mysteron Burst) and generally seems to work in most other modes ( I would have to check individually, but it is my go-to card since it used to be much cheaper than the SanDisk option whether it is now or not). BUT the Delkin does not span the new Sedna, Orion, etc, etc.

At this point, those are the two brands and those are the cards we have the most test data on for the new patches with consistent performance.

Once again - the bitrate is NOT the only factor. If the max frame size is too large, it can also cause problems on slower cards.

I repeat, if you want reliable performance for the newest patches, makes sure to get a 95MB/s card (SanDisk 64GB if you want a better shot at spanning).

Also, not all Class 10 cards are equal. The less common brand mentioned seems likely to be a low performer but I am not sure. The SanDisk Class 10 cards are faster than usual, for the most part. Hope that helps.

@rigs then you are doing something wrong. As my reel can attest, the GH2 is great with noise, when used properly. OR you have a bad sensor. (it happens, which is why I recommend people not hack their cams out of the box. That they shoot with it a bit first to see if they have a bad chip, and if so, send it in for warranty replacement until they get a good one before hacking it.)

@thepalalias Thanks very much for that informative post, it did indeed was more helpful. However like I mentioned before, I have lowered the bitrates to 50M and 100M to avoid this sort of problem but it seems it didnt help. I really like this patch as I'm all for quality but I definately knew I wouldnt be able to handle nor do I have space for 157M (if I'm correct).

When you mentioned that "the newest patches have not been designed for those cards", do you mean specifically my 7DayShop 32GB Class 10 card? or due to it's speed? Because after abit of research into my card, it has a writing speed of 20Mb/s apprantly.

So I would like to ask, is this happening to me mainly due to the writing speed of my cards? Or, the new patches just like Sandisk and more pricier SD cards??? :D If it's the former, what is the speed for a card to record video onto? :)