[Alan Okey]

You should also be aware that Compressor is not the best MPEG-2 encoder out there. For superior MPEG-2 encoding on the Mac, check out BitVice. It is slower than Compressor, but the encoding quality is much better.

https://www.innobits.com

If you compress your audio as AC3 @ 192kbps and compress the video using BitVice dual-pass VBR encoding with a ceiling of 8.3mbps, your results should be better than what you’re currently seeing.

BitVice has a demo version, so you can try it out on a sample of your video before you buy.

MainConcept is another good MPEG-2 compression program that is better than Compressor:

https://www.mainconcept.com/site/?id=769

[Alan Okey]

Just an academic observation: the Uncompressed codec will also turn any 4:4:4 RGB graphics into 4:2:2 YUV footage, so to be precise, the Uncompressed codec does throw out half of the color resolution when compared with original 4:4:4 RGB files.

[Alan Okey]

[ktofmd] “if I am not monitoring (I do not CURRENTLY have a video monitor hooked up to the system) the external drive SHOULD work? or is it safer/better to use the striped internal drive configuration?”

If you are not monitoring out through a Firewire device, the external striped Firewire 800 drive should work adequately. An internal striped SATA set will probably have a higher peak bandwidth than the external drive, but that probably won’t be an issue for SD, HDV or DVCPRO HD projects.

For uncompressed HD projects, I’d recommend at least 3 internal SATA drives striped together – and that may not even be enough bandwidth. Ideally, uncompressed HD work demands an external Fibre Channel or SCSI RAID capable of 230MB/sec. minimum throughput.

[Alan Okey]

I suspect your drive configuration is less than optimal.

When you say your external Firewire drive is mirrored, does that mean that it is two drives in an external enclosure set up as RAID 1, or that it is a single external drive that mirrors your internal OS drive?

Either way, forget the mirroring. It isn’t helping your video throughput any, and in fact may be a limiting factor. If you’re worried about data redundancy, drives are cheap – you can always back up projects and media to another drive. Instead, consider striping two internal drives (RAID 0).

All of the Firewire ports on the Mac Pro (both 400 and 800) share the same bus, so if you’re using a Firewire device for monitoring video output while simultaneously using external Firewire drives for video storage, you’re creating a recipe for potential bus contention.

Two dedicated internal SATA drives striped together (RAID 0) should provide adequate disk bandwidth to prevent dropped frames, unless you’re doing full 29.97 1080i uncompressed HD. Format the RAID 0 as Mac OS extended (HFS+), NOT journaled. Journaled file systems are really only useful for the OS drive.

[Alan Okey]

Would working with image sequences (targas or tiffs) be a possible workaround to bypass this issue?

[Alan Okey]

Welcome to Quicktime gamma hell. Macs have a native gamma of 1.8, PCs have a native gamma of 2.2. Quicktime automatically assumes a screen gamma of 1.8 on Macs, and adjusts playback levels accordingly. It’s an infuriating “feature” that can’t currently be turned off/defeated. Search the forums for “Quicktime gamma” and enjoy the fun. Let’s hope that Apple finally addresses this issue in the next release of Quicktime. 1.8 gamma is SO last century, and SO print-centric. Come on Apple, get with the program already.

There’s a program called Quicktime Gamma Stripper that supposedly solves the problem, but it only runs on Windows – go figure. Here’s the link:

https://support.franticfilms.com/wb/default.asp?action=9&fid=14&read=74

Good luck!

[Alan Okey]

https://anandtech.com/mac/showdoc.aspx?i=2816&p=16

[Alan Okey]

[THOMASW] “Alan:

Just to make sure I understand your suggestion. By “Quicktime Refernce Movie” do you mean the first export option in FCP – i.e. “Quicktime Movie” rather than compressor or quicktime converson? I assume the settings would be same as the SD sequence settings.

…then I would bring the resulting Quicktime movie into Compressor?

Can you explain to me why you believe this would result in a higher quality MPEG2 file — rather than compressing directly from the HDV sequence – which you say is the best of the options I outlined?

Thanks, Tom”

Tom,

You are correct – creating a Quicktime reference movie is done by selecting the first output option and making sure that “make self-contained” is not checked.

If you use Compressor to compress your MPEG-2 files, then there is no need to export a reference movie. I suggested the reference movie option in case you wanted to use something other than Compressor for MPEG-2 encoding, like BitVice. Sorry if I wasn’t clear.

[Alan Okey]

[THOMASW] “What is the best way to arrive at a standard definition MPEG2 file that can be used as an asset in DVD Studio Pro?

Here are options I have tried and/or am considering.

#1: Use compresser to compress the HDV sequence into an MPEG2 file. I’ve tried “90 minute, 2 pass (best quality) but the result looks a bit soft to me… softer than, say, a standard defintion commercial DVD. So… I’m trying other routes..

Compressor is far from the best MPEG-2 compression software available. Look into BitVice. If money is no object and you have a PC, the Cinemacraft encoder is the best software encoder out there, but it’s $2K.

[THOMASW] #2: Use print-to-video to make an HDV tape of the program, then use use my Canon XLHI camera to make a downcoverted (SP) dub of the program, then digitize the standard def tape into a standard def FCP project, then use compressor to make the MPEG2.

ABSOLUTELY NOT. If you go from HDV to DV, you’ll be adding an additonal stage of compression. If you dub to an analog SD format, you’ll be introducing generation loss.

[THOMASW] #3: This next option seems like it should work, but I’m not getting an image on tape. Downconvert within FCP by using Print-to-Video to make a 720/480 tape of the program. Then digitize the standard def tape into a standard def FCP project, then use compressor to make the MPEG2.

Again, this needlessly adds another stage of compression to the process.

[THOMASW] #4 Finally, I do have a Kona Card inside a G5 with dual processors. Is it an option to use the card to make the downcovert?”

Depends on which card.

Here’s what I would recommend:

Create a new sequence in FCP with the following settings:

29.97fps
720×480
16:9 anamorphic
Uncompressed 8-bit 4:2:2

Drop your HDV sequence into the new SD sequence and render. Output the file as a Quicktime reference movie, using the SD sequence settings (not Quicktime conversion).

Open the Quicktime file in whatever compression software you end up using, and make sure that you select 16:9 anamorphic.

[Alan Okey]

[Denworks] “The most noticable problem is that all straight edges look wavey. This is after I deinterlace the video in Final Cut Pro using the deinterlace filter, accounting for the appropreate field dominance.

What is the most effective way to Deinterlace my video before I compress it for the web? Can it be done on capture?”

There are expensive hardware processors that can do deinterlacing on the fly, but there’s no software deinterlacers to my knowledge that work in realtime on capture in FCP.

I would suggest looking into FieldsKit from RE:Vision Effects. It is far superior to the basic deinterlacing filter that is in FCP.

[Denworks] “One of my clients had a quicktime file that was 29.97 fps uncompressed 10-bit that appeared to be progessive. How is this aceived?”

Hard to say without more information. Do you know for sure that it was shot interlaced?