[Brian Alexander]

I would use the following workflow for these kinds of retiming jobs:

DVCProHD @ 59.94 -> ProRes @ 29.97 -> ProRes @ 24 (or 23.98).

This may give you a more reliable result.
Let us know if you have already resolved this issue.

Thanks.

[Brian Alexander]

Sorry for the delay. Are you still interested in troubleshooting help?

Let me know and I’ll be happy to provide you with FTP information for uploads.

Thanks.

[Brian Alexander]

This is great Craig, thank you. I placed an order for the CompressHD earlier today to go with the newly ordered 2.66 Dual Qaud, 12 Gig Ram Mac Pro. I can’t wait for either of them to get here!

I was a bit apprehensive ordering the CompressHD after using Elgato’s TurboHD but the Turbo seems to be geared a little more towards the prosumer side. I’m hoping that I don’t see the softness degradation like I do with the TurboHD.

I’m still waiting to hear good news from Bassel about Episode and CompressHD compatibility (I’ve got my fingers crossed): https://forums.creativecow.net/thread/132/857560#857560

Thanks again for taking the time to run these tests.

–
Brian Alexander
Sr Video Engineer
Freeman AVS

[Brian Alexander]

I would suggest you start reading other postings in this forum just to read them. Soon things will start making sense.

Digital video is one of those subjects where the more you learn the more you will begin to understand which questions you should be asking which will lead to more precise questions and answers and more confusion until one day all the information will build up in your brain and just before you feel like you’re going to explode you will reach your stage of enlightenment.

I have been regularly studying Digital Video for quite some time now and I’ve come to a point of confident knowledge but it does take time. I am still discovering things I did not know all the time. The way the industry is changing I doubt anyone knows everything. Just when you think you have all the answers they change all the questions.

Here are some basic answers to your questions:

Container Formats optimize video for specific applications. For instance, MPEG 2 is wrapped in a VOB format for DVD and it is wrapped in a GXF format for some broadcast media players and servers.

Compression is deep. One of the most widely used forms of compression is GOP (group of pictures). Basically, instead of every frame of video being an entire picture (like a strip of film) a GOP will use one frame, an i-frame, as reference then only encoded the moving parts of the picture of the subsequent frames (p and b frames) before creating another i-frame. This is basic information and each CODEC has its own specific ways of using GOP based compression; this is a study in and of itself.

Compression comes in many forms besides Codecs (COmpressors DECompressors): Interlacing, Chroma Subsampling, and odd pixel aspect ratios are also forms of compression (just to name a few).

FIle Formats? This is the same as a container.

Here are a few extra tidbits to get you on your way:

NTSC SD resolution is 720 x 486 analog and 720 x 480 digital. It is the same pixel count for 4:3 and 16:9.

Although 1080 HD video is displayed using square pixels at 1920 x 1080, not all HD codecs use that pixel space: some are 1440 x 1080 or 1280 x 1080!

720p provides a higher quality image than 1080i! (I’m sure some will argue with me here but at any given instance in 1/60 of a second, you will find a full image at 1280 x 720 and half image at 1920 x 540.)

An HD image in ProRes (SQ) @ 147 Mbps and h.264 @ 8 Mbps taxes the computers processor about the same. Why? h.264 requires decoding and ProRes is all i-frame based – nothing to decode.

I could go on…

Just remember, Wikipedia is your friend.

—
Brian Alexander
Sr Video Engineer
Freeman AVS

[Brian Alexander]

It’s possible that you have an option turned on (or off) in compressor that is not allowing a good encoding. It’s possible that creating your full size QT file straight from FCP is not being done correctly.

Because there are more features and settings in Compressor, there are more ways you can mess up your encode but it is pretty hard to mess it up in Compressor.

Here are a couple of Compressor tips which you may already know:
Make sure that your frame controls are turned on and output is set to progressive to deinterlace. Make sure your frame controls are active when resizing – use better or best (best takes forever, better is 98? as good).

For optimal quality you will do both of these processes as separate steps using ProRes as your intermediate codec. This will leave you with large files but the quality you get will be better than a quicktime conversion.

Suggested Workflow:
Quicktime Reference file (Export as Quicktime) -> Deinterlace -> Resize -> output to h.264. Long process but it is the best way to process video.

Makes sense?
Good luck.

[Brian Alexander]

Just to reiterate what the guys have already said, it is very important to find out what exactly they need. If they don’t know, maybe you can get a specific name and model number for the gear they’re using.

Some broadcast based media servers like a GXF format and Digital Cinema servers use MXF formats. Some media servers can transcode on import and others need a specific format.

Sorry to be so vague but these kinds of applications require precise communication and understanding.

Let us know what you find.
Thanks.

[Brian Alexander]

Have you tried using another i-frame based codec such as ProRes?

[Brian Alexander]

Clarification on my last post:

According to your logic, shouldn’t the 63 Mbps be much less quality on the SD Frame when it is the exact same quality as the HD Frame? It’s because of the frame size difference that yields the same quality between SD and HD but a much lower bit rate.

[Brian Alexander]

Hi Ed,

I disagree with your last statement. Frame size has a definite relationship to the amount of bits applied to a frame; after all a frame size of 1920 x 1080 (2,073,600) @ 29.97 fps, has a lot more data requiring a higher bit rate than a frame size of 1280 x 720 (921,600) @ 29.97, doesn’t it?

For instance, Apple ProRes HQ runs at 220 Mbps for a 1920 x 1080 frame. The exact same quality for a SD frame 720 x 486 (apple spec) runs at 63 Mbps.

According to your logic, shouldn’t the 63 Mbps be much less quality when it is the exact same quality? It’s because of the frame size that yields the same quality but a much lower bit rate.

I could start listing math containing pixels and bits per second but before I list differences showing math, would you please explain to me (or point me towards references), how you’ve determined this information.

This is basic math in my book. Please correct me if I’m wrong.

Thanks.

[Brian Alexander]

View this on a CRT or some other Video Display (DVD player connected to a plasma or LCD monitor). Do not use your computer monitor to troubleshoot interlacing issues.

If you burn your DVD or create your ts folder, you can open your movie in DVD Player and enable the deinterlace filter but you really should be looking at this with the correct monitor.

If you are creating this specifically for DVD I would suggest that you resize your HD footage to an SD frame size in Compressor then use the 2 SD sources in Final Cut to edit with. If your sequence settings are set for HD then you would essentially be scaling your SD video to an HD frame size then scaling back down to fit DVD 720 x 480. It’s a shame to be throwing all the HD quality away for a DVD (IMHO).

Let us know what you do to solve your issue.