http://www.redgiantsoftware.com/instanthd.html

There is a free demo available, and that may create a great upres of the 480/60P footage. I searched, and saw people use Streamclip for now, but I thought that this might provide better results.

I am getting my HD100 at the end of the week, and should try it out. Also, I have some 24P SD footage I shot in the store when I demo'd the HD100, so maybe I will try it on that later in the day.

Any thoughts, because I have seen some clips, and this puts the nail in the coffin for me regarding the HD200. I really don't have the money as it is for it, and I only want 60P for slow motion effects.

Hello there. I too have been looking for input from DVinfo members regarding the InstantHD program, but nobody ever gave me any feedback. I can tell you a friend of mine took some 24p footage from an XL2 and used the program, and I have to say, it was VERY impressive.

There have been other disussions here about it. See this thread and post by Ash Greyson for one:

http://www.dvinfo.net/conf/showthread.php?t=65661&highlight=InstantHD

A search will turn up others under InstantHD and Red Giant.

I would be interested in trying it out, but I am waiting for a good project to use it on.

Mike

I've tried the InstantHD trial version using my VX2000 footage against my FX1 footage. The trial footage has marks in it to make it impossible to use. I use Premiere Pro 2.0. Essentially, the result is definitely better than just "stretching" the DV footage on an HD timeline. And while it does a pretty good job, there is no mistaking the HDV footage from the DV footage- It's certainly a tool for using DV footage in an HDV project, but no one should think its a viable replacement for it.

I've tried the InstantHD trial version using my VX2000 footage against my FX1 footage. The trial footage has marks in it to make it impossible to use. I use Premiere Pro 2.0. Essentially, the result is definitely better than just "stretching" the DV footage on an HD timeline. And while it does a pretty good job, there is no mistaking the HDV footage from the DV footage- It's certainly a tool for using DV footage in an HDV project, but no one should think its a viable replacement for it.

I am using the trial now in FCP, and I can't figure out how to use the plugin. It has the option of input settings (DV 16X9 in my case), and then nothing regarding the output? How does this work?

In order for it to work properly the footage MUST be progressive! If you are upconverting VX2000 it has to be deinterlaced first which is a drop in quality. Canon XL2 24P 2:3:3:2 footage looks TERRIFIC, as does HVX 24p SD from miniDV.



ash =o)

By the way I was converting 480p to 720p...




ash =o)

I saw some XL2 24p 16x9 footage recently that had been lit properly and had been put thru uncompressed color correction. A friend dropped it into Instant HD and several of us were shocked at the quality. It looked better than some HDV footage that I have seen. Not being a huge HDV fan to begin with, I was very pleased to see the final results look as good as it did...

Being a Magic Bullet fan, I just came across this product. Does it make any difference on SD footage itself? As in, would it give me cleaner footage going from 720/480 DV to 720/480 Mpeg2?

Not sure about going to DVD.... in my case I was bumping to DVCproHD masters and D5 masters.




ash =o)

I was converting a Dolby Digital Trailer to HD for use in front of a short film I shot on HDV. I converted the original vob file to AVI uncompressed progressive with Procoder 2 and then, in After Effects 7, I upconverted the AVI to 1440 X 1080 with InstantHD and rendered to Cineform HD.

The quality is pretty good !

I then made a TS with the original 5.1 AC3 audio and dumped it to D-VHS.

It looks great on my 47 inch TV !

Another test I made, and it wasn't very good, was to take an old Digital 8 widescreen tape made with a cheap Sony camcorder. Since the 16X9 source was pretty bad (Sony used a digital stretching method to get the 16X9...) the upconverted footage looks horrible.

I have a 24p DV tape shot with a Panasonic DVX100: I'll try it soon and post the results.

But so far, it's worth the 100$: great little tool !

DVX does not look as good as XL2 because of the non-native 16:9, still looks pretty good though. I am telling you guys that all this "future-proof" stuff is crap because there will be more and more tools to upconvert SD footage to HD as the need arises... shooting progrssive and 16:9 might not be a bad idea though =o)


ash =o)

After seeing a demo at the local FCP user's group meeting last month, I bought a copy of Resizer (http://www.digitalanarchy.com/resizer/resizer_main.html). The reviewer stated he had used both Instant HD and Resizer. He felt the latter produced a better quality output but at a slightly higher price tag.

I can tell you that the company support for Resizer is excellent. They are a small company and real humans answer the phone and if they can't answer right there, an email comes in from someone else explaining things.

I'm not personally recommending one over the other, just making folks aware of possible alternatives.


-gb-

Resizer is best for interlaced footage no doubt... with progressive footage, results are similar in both...


ash =o)

I am telling you guys that all this "future-proof" stuff is crap because there will be more and more tools to upconvert SD footage to HD as the need arises...

It's all this upconverting talk which is nonsense as far as I'm concerned, because it's logically impossible to create HD resolution detail from an SD video stream. Sure, you can make good SD video look a little better by upsampling it carefully, but you'll never be able to add in detail which was missing in the original image. I've said many times that to prove the quality of upsampling tools all you have to do is shoot the same scene side-by-side on SD and HD cameras and then compare the SD upampled results to the HD source, but for some odd reason I've yet to see this done by anyone advocating upsampling tools. It's a simple enough experiment; you'd think someone would have done it by now...

Although I am not a professional reviewer I have tried this product.
It behaves pretty much like an old plug-in VideoPics for Premiere 6.x where an image is composed from 3 or 4 frames to give a higher resolution one. Obviously this is a very time consuming process.
The results are comparable to VHS -> S-VHS but when you compare S-VHS to DV you can plainly see the differences.
You will not get HDV quality output from any product of this type. As Kevin says you can not add what is not there to begin with.

I think this has a use when it comes to mixing HD and SD footage on the same timeline but other than that I don't see it as a viable alternative to getting the real deal. As Kevin stated, you can't add detail when non existed even if you are pulling form 4-6 frames.


Chris Watson
Watson Videography

It's all this upconverting talk which is nonsense as far as I'm concerned, because it's logically impossible to create HD resolution detail from an SD video stream. Sure, you can make good SD video look a little better by upsampling it carefully, but you'll never be able to add in detail which was missing in the original image. I've said many times that to prove the quality of upsampling tools all you have to do is shoot the same scene side-by-side on SD and HD cameras and then compare the SD upampled results to the HD source, but for some odd reason I've yet to see this done by anyone advocating upsampling tools. It's a simple enough experiment; you'd think someone would have done it by now...


It had been done, check the net for the doc "Smile" they shot scenes with an SDX900 and when upconverted it looked as good as the HD material according to the ENGINEERS! I am not advocating shooting SD forever, my point is that the only way to future proof your work is to make it compelling and well produced, that will matter 1000 times more than the resolution.

Your premise is incorrect as well... the HVX has a 540 line CCD, which is less than the PAL XL2, so... I guess it is not HD because all it is doing is shifting pixels to get more resolution.

While the high end unconvert tools are still in the thousands and tens of thousands I expect them to trickle down to the consumer soon enough. I have seen some 16:9 SD DVCpro stuff running thru a Teranex Mini that looked as good as most the HD I see on cable...



ash =o)

[...] You will not get HDV quality output from any product of this type. As Kevin says you can not add what is not there to begin with. Yes, for the most part, You're right. Some sharpening and anti-aliasing can help, but there are no miracles to be had.

I've got mixed feelings about Instant HD (http://www.redgiantsoftware.com/instanthd.html), while I think it does a nice job with anti-aliasing and presents a clean interface with convenient resizing pre-sets and sliders for sharpness, quality and anti-aliasing, I think you can achieve reasonable results with simply scaling (and maybe a little sharpening) especially if you don't have problems with jaggies to start with. SD footage that's been well shot, especially when it originated as progressive without excessive "detail" and no inter-line blurring (what's called Vert. Freq. "Thin" on the DVX100) can scale nicely to HD. Compare these images:

(1) InstantHD.jpg (http://kino-eye.com/wp-content/uploads/2006/06/InstantHD.jpg) (Upscaled w/ Instant HD, JPEG from DVCPRO HD master, 16:9 flat, 1280×720)
(2) ScaleFCP.jpg (http://kino-eye.com/wp-content/uploads/2006/06/ScaleFCP.jpg) (Upscaled w/ Final Cut Pro, JPEG from DVCPRO HD master, 16:9 flat, 1280×720)
(3) Original.jpg (http://kino-eye.com/wp-content/uploads/2006/06/Original.jpg) (Original DV progressive video, JPEG from DV master, 16:9 squeezed, 720x480).

Hmm, it's really hard to tell the difference, except in terms of render time. To be fair, Red Giant's demo footage shows off what they can do better, as it has lots of thin diagonal lines in the image. I wrote a review of Instant HD at: http://kino-eye.com/2006/06/18/instant-hd/

It's all this upconverting talk which is nonsense as far as I'm concerned, because it's logically impossible to create HD resolution detail from an SD video stream [...] Hmm, technically speaking, it's not really nonsense. If you have camera movement, sophisticated algorithms exist that can look at the previous and subsequent frames and extract additional image information from those frames and create an upconverted frame that is actually sharper and shows more detail than each original frame. I'm not an image processing expert, but I've seen several demonstrations of these techniques.

One example is DTS Digital Images (http://www.dts.com/digital_images/) (formerly known as Lowry Digital Images and was acquired by DTS last year) in Burbank California. They are a service bureau that offers digital restoration and image enhancement services. The comany was founded by John Lowry, who has developed a wide range of proprietary image processing algorithms for noise reduction, image enhancement, image restoration, and upconversion. For example, they did the upconversion from HD (footage shot with a pair of Cine Alta HD cameras and special wide angle lenses for underwater 3D) to 15perf/70mm IMAX for James Cameron's "Aliens of the Deep." There was no way an IMAX camera and all that film was going to fit in their undewater crafts, so they went with the Cine Alta HD cameras.

If you have seen "Aliens of the Deep" in an IMAX theatre, you know what I'm talking about, and if not, this is an IMAX worth trying to see for reasons of both the tech and the subject matter. Other than the typical video higlight problems, you'll see some amazing high-resolution frames, and I'm sure you'll be convinced that talk about upconverting is far from nonsense: the technology exists and over time might trickle down to the desktop. True, it's not cheap, the algorithms require tremendous amounts of computing power, but only a couple of decades ago it was impossible to edit video on the computers of the day.

Another company doing similar work but with still images is Salient Stills (http://www.salientstills.com/) in Boston, Massachusetts, they have technology that takes video frames and can make clean, high resolution images for use in print publications or in security applciations to identify subjects in the frame. The company was founded by Laura Teodosio using technology she developed as graduate student at the MIT Media Lab called "Salient Stills."

All the cool image ehancement stuff you've seen in movies like Blade Runner is more than science fiction, it's real technology we can someday hope to see trickle down into end-user applcations. It's all a matter of time.