[Jeff Pulera]

The number next to the profile name is usually the data rate, higher number = higher data rate = less compression = more quality. However, if for example you are starting out with AVCHD footage recorded at a measly 24mbps, do you need 220mbps in the intermediate file? Maybe the 150mbps version is sufficient based on the source you are using. There is a point of overkill.

Depending on Premiere version and such, the preset naming may differ, for instance on my version the DNxHD options are listed as SQ, HQ, and HQX, assuming Standard Quality, High Quality, and Extra-High-Quality.

Again, if recording some kind of 10-bit high-bitrate RAW or ProRes file in camera, then you may want to stick with an HQ or HQX (or high-bitrate) option, but if source is AVCHD or similar highly-compressed source, then maybe SQ is “good as it gets” in that case and using a higher quality intermediate will only bloat file sizes and not improve quality. Make a few test clips at different settings and check for yourself.

Whenever exporting, make certain that the Frame Size, Frame Rate, Fields (Upper, Lower, or Progressive), and Pixel Aspect (usually 1.0 for HD) match the source video. At top of Export Settings panel, you can see the SOURCE and OUTPUT settings grouped together, do a quick visual check that specs match.

Thanks

Jeff

[Jeff Pulera]

RAID 0 (2 or more drives) – STRIPED – all drives are combined for performance, full capacity available, no redundancy. If one drive fails, ALL DATA LOST.

RAID 1 (2 or more drives) – MIRRORED – same data recorded to BOTH drives so always backed up, but capacity halved. Not as fast as RAID 0.

RAID 5 (4 or more drives) – REDUNDANT – capacity of one drive in the RAID is sacrificed for redundancy. In a 16TB unit (4x4TB), you’d lose 4TB for redundancy, leaving 12TB usable. A little slower than RAID 0, but faster than RAID 1. Losing one drive is okay, but two drives go out and you lose all data

RAID 10 (4 or more drives) – creates two RAID 0 arrays, then splits data between them such that as long as ONE drive remains viable in each RAID pair, data can be recovered. Not as widely available on lower-cost RAID units as it might be on larger arrays.

For a “feature” film, redundancy will be very important. Longer-term project and can’t afford to lose your work, right?

RAID 0 is out, UNLESS you make complete backups to another drive religiously.

RAID 1 (with 2-drive system) will be no faster than using a single drive, so performance takes a hit.

Might want to increase the budget a bit and go for a 4-drive, RAID 5 system to get both speed and redundancy. That seems to be a sweet spot for performance and protection in the video realm.

For connectivity, does new PC have USB-C interface? Or possibly Thunderbolt 3? Either of those are great, otherwise USB 3.0 is pretty quick, just make sure drive has ports compatible with what PC offers.

Thanks

Jeff Pulera
Safe Harbor Computers

[Jeff Pulera]

You said the quality degrades when you import into Premiere, but then also talk about H.264 export. At what point do you see an issue? If in Premiere before exporting, check the Program window and make sure quality is set to FULL and not 1/2 or something. Also, very important that the Sequence Settings match the source material – if the frame size or frame rate are different than the source material, that will also affect what you see in Premiere, as well as hurting the output quality due to the conversion of those factors.

Assuming Sequence does match material, to get better quality output, consider using 2-Pass VBR and also increasing the data rate.

Most camcorders are recording at between 24 and 28mbps. Your export is 10mbps, and that is less than HALF of what the source was, so of course it may not look as good!

Thanks

Jeff Pulera
Safe Harbor Computers

[Jeff Pulera]

“results in the final episode render having a variable frame rate.”

I think there is some confusion here. Doesn’t matter what frame rates are used in the Sequence – when you Export the timeline, the exported file will ALWAYS have a FIXED FRAME RATE, being whatever the frame rate was set to in Export Settings. Meaning regardless of what different frame rates are used in the timeline clips, they are all get converted to the output frame rate at time of export.

So you have mixed clips in Sequence, being 23.976 and 29.97 clips. Edit in a 29.97 sequence, since that is ultimately what you want to end up with, then export also as 29.97, and any clips that are NOT 29.97 will be converted in the final output which is all 29.97 frames per second.

There may be methods of converting the frame rates of individual clips prior to editing them. Not anything I’ve needed to deal with as I always shoot my own footage, at the desired frame rate, so perhaps someone else can chime in with their favorite method of changing frame rate with best results, perhaps using After Effects?

Of course, when creating a Sequence to edit in, and also when exporting, make sure FIELDS are set correctly. Could be None (Progressive), or Upper, or Lower. Find out from whoever will be using final file, they should have the exact specs of what they need. If 1080i, then that is always Upper. If 1080p, that is Progressive. With SD video, could be either Upper or Lower, depends sometimes on various factors.

Thanks

Jeff Pulera
Safe Harbor Computers

[Jeff Pulera]

Hello and welcome PG! This is a good community and we will be happy to help with any questions or issues you may have.

Have a great day!

Jeff Pulera
Safe Harbor Computers

[Jeff Pulera]

“I was thinking of changing the render previews to Pro Res 422 (the master output codec) to give more to work with for color correcting. ”

If you apply any effect to a clip on the timeline, Premiere is going to apply that effect to the original clip. Meaning if you had already rendered a clip to “green” status – which would use the render codec you had specified – as soon as you add the color correction to the clip (or make any change) it will go back to yellow or red, meaning the rendered preview is no longer in use and you are back to dealing with the original source again, to which your effect will be applied.

So really, just apply your effects to the clips, and don’t even bother rendering previews (unless you must to get smooth playback to review the edit). Then when you EXPORT to a final delivery format, that export file is encoded straight from the original source to output format, cutting out the “middle man” thus saving a generation of loss by not using a “render preview” file.

Thanks

Jeff Pulera
Safe Harbor Computers

[Jeff Pulera]

This is not something I have done before, but I’m thinking that when you exported the anim from AE as H.264, it got compressed (of course) and that would introduce artifacting – noise and blockiness that maybe is not apparent to the viewer but is still there due to lossy compression. Then when you convert to GIF, the noise/compression errors in the H.264 requires more bandwidth to encode to a GIF.

In your example, the background is solid gray. But if that image were compressed to H.264 first, that “solid” gray may no longer be solid under closer inspection and now there are more “different” pixels to encode that are no longer a solid color area.

So what I’m saying is could you export from AE as an IMAGE SEQUENCE, meaning a series of uncompressed stills, such as PNG. That way the frames would be clean originals without lossy H.264 compression applied. Then you are starting with CLEAN frames that will maybe compress to GIF more cleanly resulting in a smaller GIF size? That’s what I would do, definitely skip encoding to the lossy H.264 intermediate clip and go straight to GIF.

Thanks

Jeff Pulera
Safe Harbor Computers

[Jeff Pulera]

I run PC, so do not have ProRes output capability, but here’s how I see it.

If you export 8-bit video, there are 3 color channels, being red, green and blue. EACH of those 3 channels is 8-bit, which in effect provides a 24-bit image. If you add an alpha channel, that is another 8-bit channel, so 24-bit plus 8-bit = 32-bit. So I think whatever you may have read is just worded confusingly. A 32-bit image already includes the alpha channel.

Export a short sample with what you think are the correct settings, bring that clip back into Premiere timeline and see if there is alpha transparency – put the alpha clip on V2 with something behind it on V1 to see if V1 clip “shows through”.

If you can perhaps post a screen shot of the Output Settings panel that actually shows the ProRes output options available, I or someone else here should be able to discern what is correct and offer some guidance, but I need something to look at to understand the Mac options that I don’t have available on my PC here.

Thanks

Jeff Pulera
Safe Harbor Computers

[Jeff Pulera]

Correct –

“And as I understand from you, I can set up my system to go:
Premiere (or Resolve) > TB2 > Ultrastudio 4K > HDMI > Dreamcolor
OR
Premiere (or Resolve) > TB2 > Ultrastudio 4K > SDI > BMD Smartview 4K”

[Jeff Pulera]

Hi Robin,

Just using an HDMI or SDI output device does not in itself guarantee smooth playback – if the computer is not fast enough to decode and play the video on the fly without dropping frames, the output device cannot help that. It can only play out frames as fast as they are delivered to it from the NLE.

If the laptop is not keeping up, you may need faster storage, or may get better results using proxies or intermediate clips that are less compressed than the original clips and therefore easier for the machine to play.

“Does that mean that if I went from my MBP > HDMI > Ultrastudio 4K > HDMI > HP Dreamcolor Z27x”

Important to note, the HDMI output from your computer has no part in this workflow! We want to bypass that. Premiere will feed video frames directly to the Black Magic output device via Thunderbolt. The display card in the computer is not used in any way to feed that output.

I don’t have any experience with the SmartView 4k unit. According to specs, it can be calibrated and provide nice images.

I freely admit that I have never used a properly calibrated broadcast monitor – I set things up by eyeball – but nevertheless I do appreciate the improved output provided by an output device versus graphics card.

Thanks

Jeff Pulera
Safe Harbor Computers