FX1000 vs AG-HMC150, esp. in low light?

Ron Evans · December 26th, 2008, 11:29 AM

Quote:

Originally Posted by Ken Ross

Ron, I just tried the file conversion and it still takes me 3X real time for the conversion when dragging over the icon. Granted I've only got a dual core T8300 (2.4 gig) with 4 megs of ram, but it is slow for me.

That is slow. Did you check to see if both cores were being used? About what it would take( I think it was 2.5 times) with a single core before I figured out the drag over icon routine. I encoded 1hour and 3 min file in about 40 mins on my Q9450 quad core two days ago. Transfer time from SR11 to PC using Motion Browser was about 20 mins. So total was close to realtime. I make sure source file is on one hard drive and transcoded file goes to another hard drive so that hard drives are not the bottleneck.

Ron

Ken Ross · December 26th, 2008, 01:37 PM

Ron, I rechecked and it's closer to 2X RT with both cores @ 100%.

As to the 'loss of sparkle', it doesn't surprise me given the number of encodes, decodes. But this would bug me no end since I'm always trying to extract the last drop of picture quality from these things. I think on my 60" 1080p plasma, the loss might be a bit more than a loss of sparkle.

Ron Evans · December 26th, 2008, 03:23 PM

Ken. I have to accept that playing AVCHD from the camera is better than going through many encode decode cycles eventually then playing as MPEG2 at about 4 or 22Mbps depending on whether it is SD or Bluray. Neither compares to the original at average of 16Mbps from the camera. However if one hasn't seen the original it is still a beautiful picture. For most people the difference is not noticable and compared to normal DVD's or cable television its wonderful!!!!! I will look forward to more AVCHD cams in the future. Would love Sony to make an AVCHD version of the FX1000/Z5 with hard drive instead of tape etc just like the SR11 I have. I am also sure that it will not be long before the NLE's have better smart rendering etc to maintain quality if there is only cuts involved. Should be a good year coming up. Now if only we could get some nice deinterlacing to smooth 60P in the displays it would be great.

Ron Evans

Steve Wolla · December 28th, 2008, 05:28 PM

I am not understanding something here....how can there be a loss of "sparkle", when all conversions are done digitally? You are just recording a bunch of 0's and 1's, right?

Now if this were an analogue process, or an analogue step was introduced, it would make sense. But not when the whole chain is digital, or did I miss something?

Ken Ross · December 28th, 2008, 05:48 PM

Steve, that generally happens when you transcode from one codec to another to allow the editing program to handle the clips more easily. The transcoding process is what causes a bit of a hit in quality.

Ron Evans · December 28th, 2008, 06:09 PM

I think Ken is correct. The playback from the SR11 to the Panasonic Plasma is vibrant. Blacks are inky black and whites are bright white. From the cameras directly to the Panasonic plasma the SR11 is better than the FX1 of the same scene if there is enough light. If I look at this in my Edius editing programs waveform monitor the range is clearly illegal for the SR11!!!! White is sometimes almost 120, blacks are right at 0. Colour has higher saturation than the FX1. So when I mix this with the FX1 video i have to balance this off a little, boosting the FX1 colour and toning down the SR11 a little. Output to HDV then playback or output to legal mpeg2 this range is not there and it thus appears to not sparkle as it once did. This by the way is with xvcolor turned off. I imagine the difference may be even more startling between an xvcolor playback and one that is then compressed to normal color range.
Clearly I would like to have this range in a camera of the FX1 quality/controls. Hence my desire for an AVCHD version of the Z5 since I think the AVCHD codec is a better codec than HDV and would allow capture at 1920x1080 square pixels. Would be nice if it were also 60P to add to the wish list!!! I wonder if the FX1000/Z5 HDMI output is from the sensors at 1920 x1080 P60??

Ron Evans

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

December 26th, 2008, 01:37 PM	#32
Ken Ross Trustee Join Date: Jan 2005 Location: New York Posts: 1,945	Ron, I rechecked and it's closer to 2X RT with both cores @ 100%. As to the 'loss of sparkle', it doesn't surprise me given the number of encodes, decodes. But this would bug me no end since I'm always trying to extract the last drop of picture quality from these things. I think on my 60" 1080p plasma, the loss might be a bit more than a loss of sparkle.

December 26th, 2008, 03:23 PM	#33
Ron Evans Inner Circle Join Date: Mar 2003 Location: Ottawa, Ontario, Canada Posts: 4,220	Ken. I have to accept that playing AVCHD from the camera is better than going through many encode decode cycles eventually then playing as MPEG2 at about 4 or 22Mbps depending on whether it is SD or Bluray. Neither compares to the original at average of 16Mbps from the camera. However if one hasn't seen the original it is still a beautiful picture. For most people the difference is not noticable and compared to normal DVD's or cable television its wonderful!!!!! I will look forward to more AVCHD cams in the future. Would love Sony to make an AVCHD version of the FX1000/Z5 with hard drive instead of tape etc just like the SR11 I have. I am also sure that it will not be long before the NLE's have better smart rendering etc to maintain quality if there is only cuts involved. Should be a good year coming up. Now if only we could get some nice deinterlacing to smooth 60P in the displays it would be great. Ron Evans

December 28th, 2008, 05:28 PM	#34
Steve Wolla Major Player Join Date: Nov 2005 Posts: 477	I am not understanding something here....how can there be a loss of "sparkle", when all conversions are done digitally? You are just recording a bunch of 0's and 1's, right? Now if this were an analogue process, or an analogue step was introduced, it would make sense. But not when the whole chain is digital, or did I miss something?

December 28th, 2008, 05:48 PM	#35
Ken Ross Trustee Join Date: Jan 2005 Location: New York Posts: 1,945	Steve, that generally happens when you transcode from one codec to another to allow the editing program to handle the clips more easily. The transcoding process is what causes a bit of a hit in quality.

December 28th, 2008, 06:09 PM	#36
Ron Evans Inner Circle Join Date: Mar 2003 Location: Ottawa, Ontario, Canada Posts: 4,220	I think Ken is correct. The playback from the SR11 to the Panasonic Plasma is vibrant. Blacks are inky black and whites are bright white. From the cameras directly to the Panasonic plasma the SR11 is better than the FX1 of the same scene if there is enough light. If I look at this in my Edius editing programs waveform monitor the range is clearly illegal for the SR11!!!! White is sometimes almost 120, blacks are right at 0. Colour has higher saturation than the FX1. So when I mix this with the FX1 video i have to balance this off a little, boosting the FX1 colour and toning down the SR11 a little. Output to HDV then playback or output to legal mpeg2 this range is not there and it thus appears to not sparkle as it once did. This by the way is with xvcolor turned off. I imagine the difference may be even more startling between an xvcolor playback and one that is then compressed to normal color range. Clearly I would like to have this range in a camera of the FX1 quality/controls. Hence my desire for an AVCHD version of the Z5 since I think the AVCHD codec is a better codec than HDV and would allow capture at 1920x1080 square pixels. Would be nice if it were also 60P to add to the wish list!!! I wonder if the FX1000/Z5 HDMI output is from the sensors at 1920 x1080 P60?? Ron Evans