From what I've seen & been told by CLIENTS (the important ones here!), down converted HD DOES look better rendered to SD vs SD itself. I've taken SD from my camera & down converted HD to SD & I've been told it looks better.

Yes, it is!

John, with the advent of HD, I used to fight this idea tooth and nail. To make a long story short, Douglas (Spot) provided some footage at a seminar he and SCS did in here Miami a few years back. Seeing it first-hand showed that my thinking was flawed.

I used to shoot with the Canon XL2 and got beautiful SD footage. When I moved to the EX3 and down converted to SD, the difference was like day and night. You can see a definite difference--a definite improvement in every way--in the EX3 down converted footage.

Unlike you, I have not run into any "hassles" at all. As I've said before, my experience shooting with the EX3 and using Vegas Pro 9 is no different, no harder, no more difficult than it was shooting in SD, regardless of the end product.

If you're finding it a hassle, then one can only presume there is a problem in your workflow.

"Is downconverted HD any better than SD?"

I think we may have missed an important distinction here:

If one is using an HD camcorder to shoot SD video internally:
--Yes. For a number of reasons. I have proven this to my satisfaction on a number of side-by-side stage shoots.

If one is shooting HD video and downconverting to SD in a NLE:
--No. Sometimes the results are worse, as has been demonstrated with AVCHD. Some detailed discussions less than a year old on this forum.

If you're finding it a hassle, then one can only presume there is a problem in your workflow.

Well I guess Walter Graff has a problem with his workflow as well.

And I guess the examples he provided were examples of his poor" workflow" as well.

Right Jay?

John

I guess there's a degree of subjectivity involved, but I'm convinced my downressed HD looks much better than SD.

It did 5-6 years ago when I bought a Z1 HDV camera, and since getting an EX1 three years ago it's even better. All shot, captured and edited in HD and then downressed to SD DVD (PAL) using standard Vegas templates.

What does "better" mean? For me, clearer definition, deeper colours, less bleed ...

There are seemingly endless articles about how to improve upon the down-converted HD to SD video quality http://www.dvinfo.net/forum/high-definition-video-editing-solutions/70792-comparison-nle-hd-sd-downconversion-quality.html There seem to be a lot of people who are not satisfied with the downrez from an NLE.

After doing my granddaughter's wedding video I found that the SD render from an HD timeline was entirely acceptable and viewing SD DVDs on an upconverting player into an HDTV the video quality seemed almost high definition, especially on closeups. Mind you I was using an HDV project and viewing the DVD on a 720p 32 inch LCD panel. A larger, 1080p panel might show more problems but my family members were very happy with the video quality.

I use a consumer model Canon HV20 in the "Cine Mode" and I would never go back to shooting DV.

The problem with downconvert has to do with video material that has a lot of motion... most video material does not have enough fast action movement to create a problem.

Perhaps it depends on apples and oranges (cameras you use). My own experience is that material shot and edited in HD and then converted for output to DVD is superior to shooting in SD. But my entry to video was HD and the SD standard I'm comparing is prosumer (not 2/3) and I've always transcoded to cineform intermediates for post. Down converting in camera is bad by report, but I've never done that in my EX1 or FX1. Clients report that they see superior image quality, whatever that means in this context.

I downconvert HD to SD all the time and it looks great. Some of what I do is very high motion automobile racing. Also, SD cameras add a fair amount of sharpening, HD cameras don't.

(By the way, in reference to another current thread, that would be "fewer hassles" not "less hassles.")

John, I cannot speak to Walter Graff's workflow or yours. But I can say that after seven years, the equipment used (both hard and soft) render the information in the article outdated.

All I can say is I know what I have seen, and continue to see, with my own eyes, and based on my experiences, I stand by my original comments.

There are seemingly endless articles about how to improve upon the down-converted HD to SD video quality
====================
Yes, a very common problem discussed on several forums over at DV Info on a regular basis. I have tried the Virtual Dub resize using Lanczos3 then frameserving to another app, and it does make a difference. It is however an example of the extra "hassles" I was referring to.

In fairness, I pretty much only shoot stage productions so the conditions are almost always poor for video. Contrasty lighting, dim lighting, awful gels (red gels, red backdrop, and dancers with red dresses on with a strong bright white light on their faces from the side). I also like to occasionally use fast motion pans for musicals. Those definitely work better with the VX2000 than with the EX1. I do love the EX1 for Hi Def video though. Truly amazing Blu-ray and HD web video.

John

John, I cannot speak to Walter Graff's workflow or yours. But I can say that after seven years, the equipment used (both hard and soft) render the information in the article outdated.
===============
It's not the first time you have made the comment about "poor workflow" Jay. You probably don't mean it in a negative way, but it sure comes across that way. How would you take it if I said that perhaps you have lower standards? Would you take that as a constructive comment? I don't feel that way so you can strike that. I did say that the article was dated, but he is not alone as others have pointed out. And some of the work he gets is pretty impressive.

John

John, I apologize for any misunderstanding.

I did not and was not saying you had "lower standards." Workflow has nothing to do with "standards" in the traditional sense of the word. Granted, the outcome may produce product of a lower standard.

If all other things are equal, then what, other than workflow, can one attribute a problem to?

"Would you take that as a constructive comment?"

Yes, I would, and have. I can't tell you many times in similar discussions with people that once all other variables were eliminated the only thing remaining was an obvious problem in workflow. Once that is resolved, the results improve.

The majority of the responders in this thread agree, based on their experiences, that HD provides better SD images than shooting in SD. The remaining replies did not answer the question directly one way or the other.

Also, I am very familiar with Walter's work and have talked with him on occasion. What I said was in no way diminishing him or his work. That would never happen.

Again, I am sorry you misinterpreted what I was saying.

FWIW:

This may be a bit of a "special case", as I'm using an EX1/nanoFlash combo. But it might interesting to some of you guys:

- the best SD video I ever produced (please account for my limited experience) has been the 50 Mbps IMX recorded by the nanoFlash.

The beauty of this method is that you can have both the HD (on SxS) and SD (on the nanoFlash CF) at the same time, shot simultaneously - it's enough to set your camera's SDI output to SD, and pick the appropriate format in the nano.

I use this method for all projects that I'm sure their SD version will be more important than HD. Really great results, and efficient workflow!

Piotr

PS BTW, another trick that you may use with this combo, when you recorded HD on both the SxS and the nano - only to discover later on that what the client really needs is the SD version:

- set your EX's SDI output to SD, your nanoFlash to IMX, and play back your HD material in camera while recording to the nano.

The idea is to avoid software down-rezzing.

--No. Sometimes the results are worse, as has been demonstrated with AVCHD. Some detailed discussions less than a year old on this forum.

I'd say just don't use AVCHD. :D While I love the idea of a tapeless format, I haven't heard any reason to use it except for that: it's slower, higher compression, etc. vs other HD capture methods.

Heck, to get decent use out of it people wanted to buy a hardware encoder for it (GPU encoding)!

"Well I guess Walter Graff has a problem with his workflow as well. "

Well for starters Walter is talking about SD shot on "broadcast" quality SD cameras and I'd agree with him. Take a $50K 2/3" DB camera with a $20K lens on it and even upconverted to HD it'll look better than most prosummer HD cameras. A fairer comparison would be a $200K HD camera up against a $50K SD camera because if you want GOOD HD that's what it should cost, 4 times what a SD camera costs.

In your case your comparing an EX1 with a VX2000 and the price difference is around 50% and something has to give. Just to get the same size photosites you'd need 2/3" CCDs.

To directly address your workflow I have never understood the obsession with using VDub or AVISynth. From my investigation Vegas uses the Precise Bicubic algorithm for scaling and it is the best there is. So why jump through hoops, some people seem to believe if it's arcane it must be better.

One thing I have noticed about HD, at least on prosummer cameras like the EX1 is it does not seem to handle blown out highlights very well and definately not on wide shots and that's a big issue with stage productions. I'd say most of my stage video is technically under exposed. No 100% zebras on anything generally. Skin just kissing 60% on zebra 2.
Now I should point out that I have dark skinned talent dressed in generally deep coloured costumes. The end result looks rich and filmic. If do get some highlights going over 100% I do not clip these in post, I pull them down to 100%.
One other suggestion. I feel the lens on the EX1 is not so good at the extreme ends of its focal length. If you make an effort to get your camera closer to the stage you might get better results.

You've said the HD from your EX1 looks great, it just looks sad when converted to SD. I'll call BS on that :) I believe you're making a very common mistake, your looking at the wrong field of view.

Try this simple test. Setup Vegas for external preview on a 24" monitor. Put your SD up and get your eyes up close so your field of view give you a comfortable image size, you're going to be pretty close the monitor. Now switch to your HD, keep your eyes at the same distance. How great does your EX1's image look that close up?

The problem is we overlook that resolution scales. This is why many people say upscaled SD looks about as good as HD. Typical SD TVs at typical vewing distances give an image at close to the limit of human visual resolution. Making it any better doesn't help, the extra resolution of HD is plain wasted.

Let me put this another way around.
Take a cheap HD camera, say good for 600 lines V res. 24" screen, it's still got to look better than SD which is limited to 400 lines. So you think hey, my HD looks OK, why does the SD look so bad. Here's what really happens. You scale that HD down from 1080 lines to 480 lines BUT you get an actual resolution of 600 * (480 / 1080) = 266 lines. Now you put that back onto the 24" screen so it fills it and it is not going to look so hot compared to a VX2000.

Now I know the EX1 is capable of much better than 600 line resolution. That's under ideal conditions with res charts. It's quite easy to get way less than that due to the lens and other factors.

Bob.

as ever bob has the technical answer at his finger-tips.... thanks.

my experience is 2 fold:

a. having shot a lot of fast action out door activity (rodeos, cross-country, etc.,) on the same camera in both hd and sd, i can assure anyone interested that the dvd produced from the hd looked BETTER than that created from sd. (all dvd were produced and created with vegas / arch)

b. i've been shooting for a company specialising in online video (with extremely high standards) for over 5 years now - they started out acquiring in sd and the last 2 years moved over to hd. the difference in quality on the web is chalk and cheese.

since i have to supply them tape, i tested out the 'practicalities' by shooting hd to tape and sd to cf and then comparing them - by no means an empirical / objective test - burning both to dvd (again, all within vegas) and viewing results on my sony hr crt, and bravia lcd. i was left in no doubt that the hd acquired looked better than the sd.

bear in mind my shots were all out door in good light, i have no idea what the results would have looked like under less optimal conditions. ymmv....

To directly address your workflow I have never understood the obsession with using VDub or AVISynth. From my investigation Vegas uses the Precise Bicubic algorithm for scaling and it is the best there is. So why jump through hoops, some people seem to believe if it's arcane it must be better.
-------------------------------------------------------
The results look better (Perrone Ford seemed to con firm this as well), but I don't actually bother because the extra steps aren't that much better. I am frameserving from Vegas to Procoder 3 which is doing the scaling. Sharpening (if any) is done by Prododer 3 as well after downscaling.
-------------------------------------------------------
One other suggestion. I feel the lens on the EX1 is not so good at the extreme ends of its focal length. If you make an effort to get your camera closer to the stage you might get better results.
-------------------------------------------------------
You are 100% right about that, but they decide where I'll be with the camera and it's usually as far to the rear of the theater as possible unfortunately. Facial details suffer the worst at that distance.

I am finding that the following EX1 Picture Profile recipe works the best so far for shadow detail when trying to tame blown out highlights:

Oliver Neubert's PP Recipe
--------------------------------------------------------
You scale that HD down from 1080 lines to 480 lines BUT you get an actual resolution of 600 * (480 / 1080) = 266 lines.
-------------------------------------------------------
Thanks Bob,

I never realized or thought about that.


John

the best SD video I ever produced (please account for my limited experience) has been the 50 Mbps IMX recorded by the nanoFlash.
-----------------
Thanks Piotr,

That may be a compelling argument for me to spend the additional cash on one. They seem to be very popular with EX1(r) and EX3 owners. Nice and small too.

John

I think I've read all these posts (but didn't click all the external links) but one thing I haven't seen mentioned may be significant for those of us in NTSC land.

Assuming the Std Def result we are talking about is for DVD, then shooting HiDef (i.e. HDV? or better) and converting to Std Def MPEG2 for DVD avoids having the 4:1:1 color space of NTSC DV.

Shooting std def 720x480 DV is 4:1:1 and must get converted to 720x480 4:2:0 for DVD.

HDV, with approx 4x as many pixels and 4:2:0 color is almost -- in a way, sort of -- like 4:4:4 color in the std def realm. Each 2x2 block of pixels roughly corresponds to 1 pixel in std def -- though interlacing has some effect there, I suppose. So the std def MPEG2 result from Hi Def should theoretically be a little "purer" and closer to the original.

Best test is using a good miniDV / HDV capable camera and recording the same scene in DV mode, then in HDV. Everything would be the same (lens, imager, etc) except the intermediate recording format.

Count me as another person here who is getting way better looking SD DVDs since I switched to HDV. No slight difference either. My HDV shot and edited SD DVDs are way better looking than what I used to get when I shot SD.

My HDV shot and edited SD DVDs are way better looking than what I used to get when I shot SD.
--------------

What kind of SD camera were you shooting with Laurence?

John