Good Results Post

[aljimenez]

This Forum is full of problems, as most forums of this type are. I seldom see posts that share how well things work, so I decided to make one.

I recently hacked my Panasonic LUMIX GH2 to make videos with very high bit-rates. I made AVCHD 1920x1080 24p with over 60Mbps and 1280x720 60p with over 44Mbps video clips with VERY high quality. Well I just got through placing them on the timeline and made different types of renderings (using fddshow codec with my own templates) and they all worked very well (pretty fast, I might add, but then look at my pc specs).

Basically, the latest Video Studio really rocks!
Al

[elijahlucian]

doesn't the GH2 already record hd video? or do you mean you are streaming it into your computer at that rate?

[aljimenez]

doesn't the GH2 already record hd video? or do you mean you are streaming it into your computer at that rate?

GH2 does record AVCHD HD video. However the AVCHD HD video is compressed, and the GH2 comes with default maximum 24Mbps. The Hack changes the camera so that it records at higher bit-rates, and thus HD video is recorded with much higher quality. Just because a camera says it records HD video it does not mean it is high quality video. All cameras record video in some sort of compressed format and the quality comes from how many bits are used to record the video... Al

[erdna]

How do you veryfy/compare the quality differences, e.g. the 24Mb/s vs the 60mb/s?

[aljimenez]

How do you veryfy/compare the quality differences, e.g. the 24Mb/s vs the 60mb/s?

Frankly, if the end of video clips is a DVD, or even worse, a Youtube posting, then recording to higher bit-rates may not be worthwhile. However, if you have a media player and high quality display device like a pc or large TV, and want to be as future proof as possible, then recording with as high a quality as possible may be what one wants.

I think it is arguable that uncompressed video would be ultimate quality (these systems are used by Hollywood film makers), and a tiny old mobile phone video is probably lower quality. So where is the "middle" ground for you? Only each of us can decide by experimenting, but for sure the direction of quality is up not down. Many bright-light video clips with not much movement and without many colors require very little Mbps to record with high quality. Low-light video clips of moving subjects with rapidly changing backgrounds will require high bit-rates to record with high quality. So, the answer to your question is that I want to lower disappointing recordings of poor quality videos. Doing a technical side by side comparison to prove definitively that higher bit-rates lead to higher quality is extremely difficult to do, and I read a lot to be convinced that this is true. Just as important is a good camera with a good encoder, and probably most importantly, the technique and composition of the actual subject matter being recorded... Al

[Shim]

Thanks for the insight = Corel supports Dell?

Never owned a dell pc or any other “off the shelf“package, so your insight into the corel/dell partnership is very interesting.

An X2 user looking for big improvements with X5 as I`m utilising far too many free 3rd party software’s to encode with X2.

[erdna]

Thanks Al for yr reply. I thought you compared on specific artifacts. Have you any idea what the GH1/2 hack exactly does? If it just multiplies the bitrate by interpollation of the existing original captured/compressed data then it will not change much. The highest PQ improvement I get is by filming in 1080p50 mode(Europe) with my Pana TM700 at 28Mb/s, edit in fullhd 50p and playback via my MM player connected to my 50" plasma TV via HDMI. Fast motion is much more natural (fluent), and eye tracked objects show up much cleaner.

[aljimenez]

I believe Panasonic has a VERY good codec technology, it may be the best.

I recall back in the late 80's/early 90's being in a company that was developing video compression chips back when mpeg1, low-res ruled the world. Truly outstanding compression was accomplished with multiple passes of the uncompressed recording. These new chips were doing live compression, while capturing. This was and is very difficult to do: namely, compress while recording. We now take this for granted, but this is very difficult to do indeed. Think about these under $1K cameras recording AND compressing when about 20 years ago this was done using parallel computers with multiple passes to get high quality.

These small cameras must hold lots of frames to make correct decisions on how to finally write compressed frames to their small SD cards, all the while more frames are coming in! I get tired thinking about how difficult this must be. I think Panasonic does the best today at this. The hacks simply change this compression to allow larger buffers, select balance of bits to motion or single frame quality, and exploiting improvements in the latest fast SD cards. The GH2 can actually write to the SD card much faster, but requires faster SD cards. Panasonic designed to slower SD cards and to stick to the AVCHD specifications. The hack removes these restrictions from the camera and it exploits how much more the camera can do... Al