BD Ripper (3D Plus) Why is software decode/encode faster than Quick Sync?

[bereid]

My question relates to this hardware and process:

H/W) Intel Core i7-2600K, Z68 chipset, 16GB DDR3, Windows 7 Pro x64
1) Rip BD media to .ISO file on local disk (whether rip is performed with DVDFab or another utility makes no perceivable difference with regard to my question).
2) Use Blu-Ray Ripper to convert locally stored .ISO file to a .MKV file on a different local disk.

My specific question (given the process above) is why the transcoding speed is only approximately 10-11 FPS when I enable Intel Quick Sync, but increases to approximately 40-45 FPS if I chose software encode/decode?

I received a DVDFab email announcement (Release Notes) within the past 2-3 months indicating (I think I'm recallnig it correctly) new support for Intel Quick Sync technology intended to speed the decode/encode process, so I was expecting when I built my new PC using the Core i7-2600K, that I should be able to take good advantage of that new feature.

Now, it is perplexing to find that by toggling only the decode/encode option between software and Quick Sync, the transcoding rate changes so dramatically in the opposite direction from what I expected. I would have expected that selecting Intel Quick Sync would take advantage of technology provided by the Core i7 that should speed the process compared with the software only option; is that an incorrect expectation? With the software decode/encode option, I have only tried the 2-pass mode, which I expect should be slower than the 1-pass option, and still, it is dramatically faster than the 1-pass Quick Sync mode.

Can anyone help me understand where my expectations may be wrong, or what other configuration options might be contributing to my puzzling performance observations?

Thanks for any advice,
Bruce Reid

[kc7457]

Intel Quick Sync (IQS) observations

My question relates to this hardware and process:

H/W) Intel Core i7-2600K, Z68 chipset, 16GB DDR3, Windows 7 Pro x64
1) Rip BD media to .ISO file on local disk (whether rip is performed with DVDFab or another utility makes no perceivable difference with regard to my question).
2) Use Blu-Ray Ripper to convert locally stored .ISO file to a .MKV file on a different local disk.

My specific question (given the process above) is why the transcoding speed is only approximately 10-11 FPS when I enable Intel Quick Sync, but increases to approximately 40-45 FPS if I chose software encode/decode?

I received a DVDFab email announcement (Release Notes) within the past 2-3 months indicating (I think I'm recallnig it correctly) new support for Intel Quick Sync technology intended to speed the decode/encode process, so I was expecting when I built my new PC using the Core i7-2600K, that I should be able to take good advantage of that new feature.

Now, it is perplexing to find that by toggling only the decode/encode option between software and Quick Sync, the transcoding rate changes so dramatically in the opposite direction from what I expected. I would have expected that selecting Intel Quick Sync would take advantage of technology provided by the Core i7 that should speed the process compared with the software only option; is that an incorrect expectation? With the software decode/encode option, I have only tried the 2-pass mode, which I expect should be slower than the 1-pass option, and still, it is dramatically faster than the 1-pass Quick Sync mode.

Can anyone help me understand where my expectations may be wrong, or what other configuration options might be contributing to my puzzling performance observations?

Thanks for any advice,
Bruce Reid

Bruce,

I too have observed inconsistent results when using IQS. Since you stated your MB is a Z68 and you have the IQS option appearing in DVDFab, I assume you are using Lucid Virtu control panel. In my case (Asus P8Z68 Deluxe MB with and overclocked i7 (4.8 GHz) and a discrete graphics card (Nvidia GeForce GTS450) I had to specifically add both the DVDFAB.exe and FabCore.exe to Lucid's application list before DVDFab would even provide me the options to use IQS.

Like you, I initially covert my personally owned Blu-rays to an .ISO file and then either use DVDFab's Blu-ray Copy - Full Disk to make a BD25 Blu-ray copy or DVDFab's Blu-ray Ripper to take either my .ISO image or the BD25 copy and convert it into a .m4v file for my iPad / iPod Touch, etc.

Your results will vary from mine, since we have different media and different systems, but for a baseline here are some recent results (run today) using DVDFab v8.1.3.8 and the Blu-ray version of The Lion King where I converted my original Blu-ray (about 30GB) to a BD25 image (using DVDFab's Blu-ray Copy - Full Disk). The whole process took about 68 minutes and the main movie (00084.M2TS) averaged 13.40 MB/s with my CPU averaging 65-70%. Naturally IQS was selected and I only used one pass and I did opt to enable Remove HD Audio.

When I used the same Lion King .ISO file to create the .m4v it too about 10minutes and averaged about 206 FPS, using on average 45% CPU.

I have notice that in previous releases not selecting Remove HD Audio during DVDFab's Blu-ray Copy - Full Disk process resulted in greater MB/s speed (about a 40-60% improvement), but this bug seems to have been fixed in more recent release of DVDFab (but you may want to experiment with that option) - as counter-intuitive as it may seem.

I wish i could be more helpful, but there are several older posts from other members asking for a tutorial on what settings to use to fully optomize very new Win7 systems that have the iCore 2 unlocked processors with the Z68 motherboards / Intel Quick Sync. Unfortunately I have not seen such advice from the DVDFab developers, so for now it is a bit of (time consuming) experimentation.

Hope this helped....

Ken