Tape Based Capture and Sync

[Matthew Riggieri]

Hey

Hitting some walls for a workflow we are developing.

Currently testing bothPAL and NTSC Hi8 Cameras
Capture via Decklink Studio At wither NTSC (525/59.94i or 625/50i) to ProRes 422HQ
Set Project up as Progressive project (NO FIELD PROCESSING CHECKED)
Import into Resolve
Subclip based on Sc/Tk
Sync to Audio

Aiming to export AVID MXF media that would match back to original interlaced TC. Everything look great in the timeline. When exported using individual clips, frame rate to either h264 or MXF export is doubled and frames are doubled.

Unsure where we are running off course.

 

[Tony Sal]

Matthew,
In my experience ProRes 422HQ is overkill for analog NTSC hi8 footage capture. ProRes LT would be acceptable. In a progressive project NTSC timecode would be double so you are seeing 59.94 timecode instead of 29.97 or 25 frame timecode. The conversion from interlace to progressive is creating frame doubling. Do you wish to remain in interlace ?

 

[Jef Huey]

We really need to know what the goal of this workflow is.

You say you are aiming to "make AVID MXF media that would match back to original interlaced TC." And you say you have set the project as Progressive.

So I am guessing that you MIGHT be wanting to take interlaced footage from DV and make it Progressive for use in a different project. If that is the case, then this intermediate project's frame rate should be 29.97, not 59.94.

But without a full understanding of what you are wanting, it is only a guess.

Another thing to look at is in the timeline settings under Format. There is a setting near the bottom labeled "Align Clips to Frame boundaries". This affects how interlaced material is shown and handled in a sequence. I am not sure if it has an impact on final render. Never tested that.

Good luck,
Jef

 

[Perry Paolantonio]

In my experience ProRes 422HQ is overkill for analog NTSC hi8 footage capture. ProRes LT would be acceptable.

There's nothing overkill about it. ProRes LT has more compression than HQ. Even in today's world of high hard drive prices, the space an SD ProRes HQ file takes up is inconsequential, so why capture to a file that has more compression and risks introducing artifacts?

 

[Tony Sal]

There's nothing overkill about it. ProRes LT has more compression than HQ. Even in today's world of high hard drive prices, the space an SD ProRes HQ file takes up is inconsequential, so why capture to a file that has more compression and risks introducing artifacts?

Given analog standard definition formats such as Hi8mm are already suffering from image quality loss I see little gained in encoding to a fat codec such as ProRes 422HQ. Garbage in Garbage out does not result in sweet smelling trash. Prores 422 would have been another suitable option.

The recent outrageous increase in media storage pricing is a major factor when weighing the cost benefit ratio for media storage vs perceived image quality gain. Attending NAB 2026 was beneficial as I was able to speak directly with a leading hard drive vendor representative who warned now is the time to purchase media storage as pricing will continue to increase in the coming months given the impact of Ai Data centers driving the cost up.

The 2026 storage crisis: Why AI data centers are hoarding every hard drive on the market

Why spinning rust won't save you from the global SSD price surge

www.howtogeek.com

Quote from article
"The situation has reached such an extreme that major storage manufacturers like Western Digital have publicly confirmed their entire hard drive production capacity is completely sold out for the entirety of calendar 2026. Data center giants have even locked in firm purchase orders for hard drives extending well into 2027 and 2028, effectively draining the supply pool for everyone else."

 

[Perry Paolantonio]

Given analog standard definition formats such as Hi8mm are already suffering from image quality loss I see little gained in encoding to a fat codec such as ProRes 422HQ. Garbage in Garbage out does not result in sweet smelling trash.

This is, IMO, 100% the wrong way to look at it.

For one thing, analog SD video may not be high quality by today's standards, but it's analog and as such, it has no compression applied to it. By capturing to LT, you're applying digital compression on top of that, potentially reducing the quality of the capture and baking in artifacts that are not part of the original footage.

The idea that "it's low quality so you shouldn't use a high quality capture format" is pervasive in the post production world, and is wrong. It's the same logic as not capturing film at high resolution because the film itself couldn't resolve image details, so you should just scan to HD to save space. That completely ignores other factors, like the side effects of upscaling from HD to 4k for display. If you capture at 4k you avoid that upscaling and the end result is markedly better.

Drive prices may be temporarily high at the moment but you're talking about an insignificant savings in disk space with SD video when you capture to LT vs HQ. SD video doesn't require particularly fast drives either.

For reference, we just completed a tape capture job for a local university library last week. 96 Umatic tapes, totaling about 32 hours of footage. The files were captured to NTSC ProRes 422HQ, and come in at a whopping 1TB. That's not exactly a lot of data, nor an expensive drive - about $100 for a 1TB USB3 drive, which is more than fast enough to play back SD video smoothly. You might wind up saving what? 200GB by going to LT? for 30+ hours of footage? It's simply not a savings worth even considering.

 

[Tony Sal]

This is, IMO, 100% the wrong way to look at it.

For one thing, analog SD video may not be high quality by today's standards, but it's analog and as such, it has no compression applied to it. By capturing to LT, you're applying digital compression on top of that, potentially reducing the quality of the capture and baking in artifacts that are not part of the original footage.

The idea that "it's low quality so you shouldn't use a high quality capture format" is pervasive in the post production world, and is wrong. It's the same logic as not capturing film at high resolution because the film itself couldn't resolve image details, so you should just scan to HD to save space. That completely ignores other factors, like the side effects of upscaling from HD to 4k for display. If you capture at 4k you avoid that upscaling and the end result is markedly better.

Drive prices may be temporarily high at the moment but you're talking about an insignificant savings in disk space with SD video when you capture to LT vs HQ. SD video doesn't require particularly fast drives either.

For reference, we just completed a tape capture job for a local university library last week. 96 Umatic tapes, totaling about 32 hours of footage. The files were captured to NTSC ProRes 422HQ, and come in at a whopping 1TB. That's not exactly a lot of data, nor an expensive drive - about $100 for a 1TB USB3 drive, which is more than fast enough to play back SD video smoothly. You might wind up saving what? 200GB by going to LT? for 30+ hours of footage? It's simply not a savings worth even considering.

Given Matt mentioned he was using a camcorder as the playback source most likely via the composite output and without the aid of a timebase corrector or stabilizer I will say that is a major problem which encoding to ProRes 422 is not going to improve.

Keep in mind transferring to one drive is just the beginning of a long term archival plan as additional backup copies will have to be made throughout the lifespan of the archival material. When it comes to digital asset management my mantra is "one is none, two is one and three is just about right." As I mentioned media storage pricing will continue to rise into 2028 so one must consider the pro and cons. However if you can afford the ever increasing media storage pricing and higher data usage is not an issue then encode at the best codec which meets you needs but for those on a budget consider the alternatives.

 

[Perry Paolantonio]

Given Matt mentioned he was using a camcorder as the playback source most likely via the composite output and without the aid of a timebase corrector or stabilizer I will say that is a major problem which encoding to ProRes 422 is not going to improve.

No, this is Hi-8 so S-video is more likely. A TBC would be needed no matter what.

That being said, it's not relevant whether it's composite or S-Video, or component. The analog video has no digital compression artifacts because it's analog. Digitizing the file introduces the possibility of adding compression artifacts to the image that were not there before. Capturing to ProRes LT will do this, therefore it's an unsuitable format to capture to. It's designed to be a decently high quality proxy file, not a master format.

Matthew doesn't specify the purpose of this project, but you bring up archival formats. That's all we do, every day. It's taken a decade to convince most (but not all) of our archive customers to use ProRes instead of Uncompressed 422. We did extensive testing on the format and determined that there is no compression loss worth worrying about, even for archival use. But that was with 4444 (we didn't publish the 422 results, but the compression tests were identical to 4444).

When you're talking about SD video, the space savings between 422HQ and LT is not significant enough to make a dent, but if you capture to that format you've added compression artifacts that are irreversible. Why do that?

ProRes LT is something we use when we upload a graded file to frame.io for a customer, and you can absolutely see compression that's not present in 422HQ. We do it because space on frame.io is at a premium (much more so than with hard drives, even now), and the color is no different between the two, just the amount of compression applied.

Just using my example above: a 32 hour capture set takes about 100GB as ProRes 422HQ. On SSDs that's maybe $150. On spinning drives, $80-100. For almost 100 tapes. If you did it on ProRes LT you might save a couple hundred GB but you'd still need to buy the same size drive to put it on. So again, why make the image worse?

 

[Rajneesh Kassin]

Matthew doesn't specify the purpose of this project, but you bring up archival formats. That's all we do, every day. It's taken a decade to convince most (but not all) of our archive customers to use ProRes instead of Uncompressed 422. We did extensive testing on the format and determined that there is no compression loss worth worrying about, even for archival use. But that was with 4444 (we didn't publish the 422 results, but the compression tests were identical to 4444).

An old friend of mine does archival videotape transfers and restoration out here on the West coast, and I can't convince him to stop using ProRes 444. We've had an argument for almost 10 years that ProRes 422 (or maybe 422HQ) would be fine for 525 SD videotapes from any source, and that includes master 2" quad videotapes. He still insists on 444, which I think is a waste of space for 8-bit standard-def video.

On the Hi8 question: we found that if you play a Hi8 take on a Digital 8 player, you can actually get a solid SD signal out of it, roughly the same as it would be with a TBC. The problem there is all of them were limited to Firewire, which is pretty outdated connector for 2026.

 

[Perry Paolantonio]

An old friend of mine does archival videotape transfers and restoration out here on the West coast, and I can't convince him to stop using ProRes 444. We've had an argument for almost 10 years that ProRes 422 (or maybe 422HQ) would be fine for 525 SD videotapes from any source, and that includes master 2" quad videotapes. He still insists on 444, which I think is a waste of space for 8-bit standard-def video.

Most standard def video is analog, there are no bits. It's not 8 bit unless it's a digital SD format.

That said, 4444 is probably unnecessary in most cases, because 422HQ already covers all of what's on the tape and then some. We use 4444 only when we're capturing a format that's 444, like HDCAM SR. Which comes up like once a year, if that.

But I still don't understand this obsession with saving space with ProRes. if you were comparing Uncompressed 10 bit to ProRes, or DPX to ProRes, or if we were talking about high resolution like 4k or more, then I get the space arguments. but the difference between 4444 and 422HQ for SD video isn't enough to make much of a dent even if you're capturing hours and hours of footage. There's just no harm in capturing to a higher quality format, on any level. So why make an argument about space savings when we're talking about trivially small files?

I mean, we all work with massive files all day long here. That's where you need to think about this. Not SD (or even HD) capture.