-
Avid VS Premiere VS FCPX VS Edius?
-
[Oliver Peters] ” The point of the Teranex units is high-quality upscaling, not 1:1 capture. If you have SD and want it in 1080, 2K, etc. the Teranex will give you superior results.”
OK, I’m willing to believe this.
However, I wish somebody would post real world captures or other observable examples of HOW MUCH improvement is gained by spending the big bucks on these processes.
I can understand that if you have a noisy VHS tape, a smart algorithm should make a difference. But if you have a CLEAN VHS tape, is the benefit as much? Half as much? 5% better?
Basically, whats the point where spending the money on an outside service makes a visible difference?
At what point does interpolating the 640×480 pixel raster onto a 2k signal via A verses B make any actual visible difference? That’s what I want somebody to test. So far, I can’t find that out – no matter how much I try.
Oh well.
Creator of XinTwo – https://www.xintwo.com
The shortest path to FCP X mastery. -
[Bill Davis] “However, I wish somebody would post real world captures or other observable examples of HOW MUCH improvement is gained by spending the big bucks on these processes.”
Well, I could show you the differences between captures from a BETA tape…if I can figure out how to capture that via firewire. I’ll have to see if I can find a j-30 deck or something. And then capture that via firewire, as DV…and upconvert with Compressor. And then capture that same tape via AJA IoXT upconverting via hardware. (Don’t have access to Terranex for a test like this). I have done tests in the past of this, and the difference is visually noticeable. I could tell that the AJA/Terranex upconvert looked better. The fact that you won’t believe me or Oliver when we tell you this is odd…don’t trust us? Don’t trust other professionals when they say this?
[Bill Davis] “I can understand that if you have a noisy VHS tape, a smart algorithm should make a difference. But if you have a CLEAN VHS tape, is the benefit as much? Half as much? 5% better?”
RARELY do I have VHS masters to work from…it’s primarily Beta, 3/4″, digibeta…either film telecined to those formats long ago, either high end means, or sketchy (projector directly into a camera lens) means. Or broadcast news masters on those formats. The difference can’t be quatified by a percentage…I don’t know how to say “it looks better…maybe 10% better than the other.” I can say that the difference is noticeable…it’s sharper, more detail. Less artifacts.
[Bill Davis] “Basically, whats the point where spending the money on an outside service makes a visible difference?”
Because my master isn’t the FINAL that airs. It gets converted to a broadcast master format, so a layer of compression. And then when broadcast, it’s compressed over the airwaves by various means…cable or satellite. And then they might make international masters, so more converting and compressing. Because of this, networks insist on having the best possible conversion available, because it’s going to be compressed and converted again..and if you start out with a better image, it’ll hold up more than one that’s OK…but not as good.
But again…this is the world of high end video broadcast…and the standards there are very high. Crappy video only gets accepted when it’s a) one of a kind video, very important to the story… and b) zero alternatives exist, it’s the only version of the video, and there’s not another source to replace it. Then networks reluctantly take it. I’ve had to deliver shows with HORRID interlacing baked into the video…and with masters that smaller local markets archives as “high end WMV” files…meaning WMV compressed at 640×480. And yes, I’ve done a machine to machine output upconvert to 1080, along with a software upconvert, and the difference is visible. So why not go with that extra 5-10% (if it’s even that quatifiable?) We are talking shows with budgets from $125,000 to $350,000 and up…full of archival.
Shane
Little Frog Post
Read my blog, Little Frog in High Def -
[Bill Davis] “There’s a long thread somewhere on the forum during the earlier debates about whether there were any visible advantages to going the Terrenex route over the direct FireWire deck transfers. “
For the curious (and the nerdy).
https://forums.creativecow.net/thread/335/85274#85363
[Bill Davis] “Basically, whats the point where spending the money on an outside service makes a visible difference?”
I think that probably always comes down to a case-by-case basis. What’s the budget, what’s the expectations, etc., Most people opt for ‘best quality possible’ which really means ‘best quality within the budget’.
-
[Bill Davis] “I can understand that if you have a noisy VHS tape, a smart algorithm should make a difference. But if you have a CLEAN VHS tape, is the benefit as much? Half as much? 5% better?”
Beauty is in the eye of the beholder. I would say subjectively that the improvement is between 5% and 25% depending on the results you are trying to achieve. Remember I’m only talking about conversion and scaling, not capture; however, the BMD Teranex models can also function for i/o.
https://digitalfilms.wordpress.com/2016/07/30/blackmagic-design-teranex-processors/
[Bill Davis] “Basically, whats the point where spending the money on an outside service makes a visible difference?”
You also need to separate the discussion between the original Teranex products and those currently made under the BMD banner. There are certain image restoration products that Teranex used to make, which are no longer part of the line-up. Although some of those functions have been integrated. For example, on an indie film that I had to piece together from rather dirty cut negative, I used a Teranex image restoration product (no longer sold) to reduce a significant amount of embedded film dirt (in real time). However, for standard use these days, the BMD products are quite good and relatively cheap.
[Bill Davis] “At what point does interpolating the 640×480 pixel raster onto a 2k signal via A verses B make any actual visible difference? “
I would contend that the bigger the target, the more obvious the difference in quality. The fact that it’s crap to begin with is not an absolute mitigating factor. There are many variables that affect that subjective 5-25% judgement range. For example, is there a lot of motion? Do you need to convert interlaced to progressive? Do you need to add noise reduction and/or sharpening?
On a recent job for a documentary about the WWII-era WASPs (forgotten female pilots), I used a BMD Teranex convertor to uprez all of the interview footage from SD to HD. These were all static shots in limbo of older ladies. Because they are on screen alone for lengths of time, you focus in on image quality. There was a noticeable difference between going the Teranex route or not. Could I achieve similar results within a few percentage point of the Teranex? Probably – but not in real-time. Think about how much of a beast it is to render Neat plug-ins. Much better going this way.
In addition, some network delivery requirement specify the use of Teranx and/or SAM (Snell) conversion products when content is from SD sources. OTOH, if you are working on a few clips for basic corporate jobs, then no, it’s probably not worth it.
– Oliver
Oliver Peters Post Production Services, LLC
Orlando, FL
http://www.oliverpeters.com -
BTW (regarding Teranex) – A great way to test these things is with a DVD and Blu-ray test series called HQV Benchmark (Google it.) It’s quite old at this point, but something I developed with the Teranex folks when they were together with Silicon Optics. At that point, they were trying to license some of their algorithms into consumer products, particularly high-end DVD players and TV sets. If you can snag one on E-bay, it includes a series of tests and grading criteria to evaluate some of what we’ve been discussing. It’s designed for consumers and TV sets, but would still work in professional video applications, too.
– Oliver
Oliver Peters Post Production Services, LLC
Orlando, FL
http://www.oliverpeters.com -
Interesting info Oliver. A little of topic but the info is more than welcomed by me.
-
-
[Bill Davis] “In the end, they dumped ALL the external box approaches including those via high-end Hollywood transfer services – and simply plugged in appropriate tape decks via Firewire and digitized directly into X.”
I asked Patrick about this, and the workflow he did for me on LOST TAPES: Pearl Harbor, and he said you were right about everything but the firewire capture directly into FCX. The original capture for OJ was SD through Media Express with BMD capture hardware. But then they scaled up in FCX..and compared that to what a company did with a Terranex and the terranex was softer.
But for Pearl Harbor, they used the AJA Kona 3 and FCP 7, and that worked great for them. (I do the same thing…still utilizing FCP 7 on occasion…or the AJA capture tool).
Shane
Little Frog Post
Read my blog, Little Frog in High Def -
Not surprised I miss-remembered that step.
What blew me away at the time was that the native algorithm in X was good enough to do the ultimate storyline to master without the need to go out of house.
(A radical idea back in an era where the default expectation was that you had to toss expensive and sophisticated hardware into the signal chain for ALL transcode broadcast deliveries.)Changed my thinking at the time.
Thanks for the clarification.
Creator of XinTwo – https://www.xintwo.com
The shortest path to FCP X mastery. -
I have no doubt that the scaler in X is good…far better than FCP 7. I know that Adobe’s is good too. And I’d rely on them for many things. The fact that they did that test and X was better doesn’t surprise me.
Shane
Little Frog Post
Read my blog, Little Frog in High Def
Reply to this Discussion! Login or Sign Up