I spent from Friday til early Sunday trying to convert 25 8-10 minute MOV clips to Cineform avi. It was a new install of Cineform and i had set Quality to High - well this aft Sunday, I stopped this madness and set quality to medium.

Getting 123,407KB/min using medium whereas the high was taking four times as long to process.

I just wonder what the difference is in the final file I import into Vegas Pro. Of course we all want the final product to look as good as possible but I don't understand the manual's reference to "lot of post-processing" as being a reason for high quality setting.

Could someone elaborate on this?

TIA
Harry

The is only 5%-10% difference in encode times between Medium and High, so something else went wrong in your testing. Medium vs High is the same compute process, only the data rate is different, this increase in data can only slightly effect performance.

So on a decent PC, what would be the data rate of conversion of a 1080p MOV conversion to 1080p AVI Cineform?

I've wondered how big a difference upgrading my box from my Athlon X2 2.1Ghz to say a i7 box etc. The conversion the way I did it was painfull though it was 24 MOVs it took forever.

My settings for conversion of my 5Dmk2 MOVs is Medium, Maintain Source frame format option selected, Progressive Source checked, iframes only not checked.

If you don't have a 'lot of post-processing', why not edit native?

Sareesh,

Starting to wonder the same thing. Initially I bought Cineform years ago because the MOVs would hardly play in the editor and I needed to apply sound and video fx and play to see how it looked/sounded. At one point the converted AVIs would also play back well in a media player but lately they play for a few seconds, stop, pick back up almost as if I'm streaming the video from the web but worse.

Are you editing MOVs from a 5dMk2 natively in Vegas Pro?

I edit natively (@ 25fps) in Premiere on my core 2 duo laptop. I did try Cineform for a short in 2008 (Was shooting HDV, monitoring via Onlocation and ingesting via firewire into Cineform), but I decided to ditch it - on my keying and comping tests, I found cineform wasn't really necessary.

If you shoot good chroma on an HDSLR, it will key well; and prores or cineform will be superflous. If you shoot bad chroma, nothing can save you except heavy image manipulation (in which case an uncompressed 16-bit image sequence is what I suggest) in 32 bit mode - and even then the results will suck.

If you don't have a 'lot of post-processing', why not edit native?

be cause Cineform is more than a simple DI codec....
- active metadata per dinamic correction without re render
- more color space for editing
- faster to play and edit than original codec
just to spend some words about cineform...

if you think that you not need cineform probably you ever used the real power of cineform.
a simple example :
cineform metadata edit by firstlight = correction in realtime for 3d, color, and more NO ADDITIONAL RENDER for all no linear editor, most of color correction needs more render time, with cineform you not have it.
Metadata correction allow you to build a database of correction in a different computer from original editor, and you can send few kb instead to pass full rendered clips...
and more...

So, back to the original question: How much quality do you lose going to medium? I have been a Cineform user since it was first released and I remember when medium quality was the recommended one and all the online quality examples (which looked great) were done at medium quality. The version of Cineform that came with Sony Vegas was locked at medium quality as well and I don't remember ever having an issue with the quality, or anyone else having an issue with the quality of that version. Medium quality is still as good as it was back then I assume.

Medium is fine for all but the biggest color correction changes. Without color correct Medium through Filmscan 2 would look the same. The subtle artifacts can show with Medium as you dig into the shadow detail or greatly post-zoom into the scene. Medium is the default for most of our tools as it is good enough of many applications.

That is amazingly helpful information. So if I am doing a lot of cropping and color correction in FirstLight, high quality is a good idea, but transcoding stock footage that already looks great will look just as good in medium quality. Cineform masters that are post color correction and cropping would look just as good in medium quality as the would in film scan. Am I getting the right idea?

Is using the film scan modes on say footage that was shot in dark conditions and will need a huge amount of brightening and noise reduction a good idea?

One more question then. It seems to me that if you are using an AVCHD camera (in my case a Nikon D5100 DSLR), that a lot of that information that would be the difference between the high and medium quality Cineform probably isn't there to begin with. That with a DSLR or an AVCHD camcorder compression algorithms you can't really color correct that heavily to begin with, and whether I choose medium or high compression is going to have very little effect on the end quality with this source material. With an F100 there would be a big difference, but not with a DSLR using 18Mbps AVCHD. Is that a correct assumption?

I realize that the small difference in storage space and overhead is probably not that big a deal to many people, but in my case where I am using an Intel Core2Duo laptop with external USB 2 drives, the difference between these two modes is a bigger deal than it would be on a more up to date system and I will take a 20% boost in performance with a negligible difference in quality if that's using medium quality would mean in my situation.

Your ideas are correct. However even with highly compressed AVCHD you can get a little more out of them with High than Medium, but the more compressed the source the diminished returns.