How much difference when edited video is re-encoded vs smart rendered?

[perfection]

Is it right that that the rendering engine will make a difference?

Let's assume a video editing program where i have introduced a title and a few transitions on the clip cuts on an mp4 video.

I run this through the smart re-encoding feature of the program which only re-encodes the manipulated portions of the video v/s rendering it with full re-encoding (not using smart render feature) - How much decrease in quality would there be and is this quantifiable (apart from trusting the naked eye)?

Would just like to understand te concept and what i am losing out on by not having a smart render feature in my editor!

Thanks experts at Videohelp

[KarMa]

Anytime you re-encode with a lossy codec, it is going to lose some quality. Even at the same bitrate. I have no way of quantifying it as it just depends on how good the codec is and what settings you used and how complex the content is (movement).

[perfection]

Ooops i did not make myself clear

I meant if i used the same codec (eg mp4 x264) and the output after a few edits was also the same (mp4 x264) then what??

[Cornucopia]

Same answer.
Smart rendering means copying that clip segment (if possible) vs re-encoding. So lossless vs lossy. Sometimes lossy isn't noticeably worse, sometimes it is. The underlying pixels ARE always degraded, though.

You could run a few tests and compare at certain bitrates using psnr or something as a quantitative objective measure, but it's always ultimately a qualitative judgement call, which is mainly subjective.

Scott

[sophisticles]

In theory the "smart rendered" version may be of somewhat higher quality, due to not having to re-encode the whole file.

Just curious, are you working with Pegasys' software?