24F Not Good Enough For Film Transfer At DVFilm?
 

I posted this over at DVXuser, and it seems to have elicited mostly shrugs, and "what the....", like myself, so far. Does anyone KNOW if DVFilm is alone in preferring 60i for film out from the H1, A1, & G1, or are Duart and DFG requesting the same? Not that it really effects that many projects for the most part, it would just be good information to have going into the purchase, or at the start of a planned HDV to Film. Anyone care to take a stab as to why Marcus is lumping Canon's 24F in with Sony's 24P, as opposed to saying that it is tomato/tomahto with Panny's 24P, as the arguments have sort led to here and at DVXuser. Who exactly IS accepting/preferring 24F for film outs?

Here's the link and important information from DVFilm....

http://www.dvfilm.com/fx1/

(I know it says FX1, but it is about the XL-H1, and I'm assuming G1, and A1 as well)

"FX1 or XL-H1 Euro model: Shoot HDV-mode 50i (interlaced) 1/50th sec shutter speed.

FX1 or XL-H1 US Model: Shoot HDV-mode 60i (interlaced) 1/60th sec shutter speed.

Z1: Shoot HDV-mode 60i (interlaced) 1/60th sec shutter speed.

It's very important that you use the recommended shutter speeds for a successful conversion to 24 or 25 fps.
Do not use Auto Exposure where the camera sets the shutter speed. Also watch the white clipping using the Zebra stripe feature. Do not allow significant white clipping particularly in scenes with camera movement, as the white clip areas will appear to have stuttered motion.
Do not use CF25, 25F, CF24 or 24F modes. These modes result in reduced vertical sharpness vs. the interlaced modes (see examples at bottom of this page). Do not use CF30 mode, it cannot be converted to 24P."

It's one thing to say that 24F and 24P are the same, but if film transfer houses are discriminating between the two, it should be amended to, they are the same, EXCEPT for film out purposes. It always seems Marcus has been ahead of the trends for the most part, seems odd that this distinction is being made for tape to film transfers, and probably should be more widely known, especially for those folks spending tons of buck to get their 35mm adapters for their A1's, what not, with the anticipation of transfering to film. Is 60i really the only/preferred way to do that with Canon's HD line?

It looks , to me at least, that the author of the article believes 24/25F from the Canon cameras is equivalent in quality to the Sony CF modes. The Canon F modes are far superior to the Sony CF modes as we know.

There was a film shot on the XL-H1 called "The Signal" it would be interesting to know what mode they shot in.

TT

Nick Hiltgen's "The Signal" has been heavily discussed here on DV Info Net. Please refer to our XL H1 forum for all of the details.

The reason for the shrugs is that most of the people on these forums have never, nor will ever do a film out. So debating 24P vs 24F, DVCProHD vs HDV, Sony vs Canon vs Panasonic is pointless.

If your planning on doing a film out start by working with the company thats going to do the film out and produce a test. The test will do several things for you, you'll get to see the quality of the output but more importantly together you will determine the EXACT workflow you need to follow to have good looking film.

Not only are there variables on the video side but also the type of film recorder, film stock, file format (some codecs lend themselves better to film outs than others) etc..

I think you'd find that no two film out facilities do it the same. So the best thing you can do is start working with the company you plan on doing the film out before you START production.

Unfortunately I've only been able to find that they recorded to HDV tape. In the Avid article they go into some depth but there is no mention of which mode they recorded in.

Any ideas Chris?

TT

Well I think the easiest thing to do is to simply ask Nick. He shot it, and he is a regular, frequently contributing member of DV Info Net.

That's why I said this in the original post....

"Not that it really effects that many projects for the most part, it would just be good information to have going into the purchase, or at the start of a planned HDV to Film."

This is most certainly not intended to spark a flame war, but I do think the lack of available information as to specifically which companies would do 24F to film transfers is worth having in the search somewhere, as right now the impression is that anything 24P can do, 24F can do just as well. Just like many DVX owners record on 24Pa for that just in case, it think it might be important, at the very least, and probably more likely for XL-H1 owners, or those who are planning on adding 35mm adapters or what not, for them to know that the "just in case" would be 60i, and that 24F is for HD projection and DVD distribution only - if that is correct, and I'm pretty much hoping it's not correct personally.

Yes, and there is an important reason for that impression, because for all practical purposes, 24F is 24P. The means may be different but the results are the same.

The argument that "well, there's a slight loss in resolution in 24F" doesn't make much sense, because even with that slight loss, 1080p24 shot in Canon Frame mode is still at a higher res than 720p24.

So, once again:

1080p24 (minus the hit in vertical res from Canon Frame mode) is greater than (or at least equal to) 720p24.

As far as the "p24" part of that is concerned, Frame mode is progressive scan. There's no difference in the motion signature, the cadence of 24F vs. 24P. They're identical.

Chuck Spaulding is quite correct in his assessment that not many shooters in this forum or other web communities are transferring their projects to 35mm film, so I fully agree that this topic is pretty much a non-issue. The main requirement for a successful 35mm transfer is a previously established, good working relationship with a transfer house, whether it's DV Film or DuArt or whoever. Shoot it the way they particularly want to receive it, and everything will be fine.

I call BS on this one. It seems like more misconceptions and assumptions regarding the format.

How is the footage from an XL2 or DVX100 any better than the reduced resolution of these Hi-Def cameras? It isn't. Especially in the case of the XL-H1, XH-G1, or XH-A1.

First, they never give any reasoning as to why they lump the XL-H1 in with the Sony cameras. Any common joe that reads these boards can tell you the frame modes are completely different. Second, the camera delivers 24 discreet frames per second. Ingest the footage properly into an NLE and it's no different than the other "24p" cameras they list. Third, in their list of deliverable formats how would they know the difference?

"File formats can be Quicktime 5(any codec) or nearly any image-file format such as JPEG, PICT, BMP, TGA, SGI or TIFF, and AIFF or WAV files for audio."

You see my point? As an example, you could deliver the film in an uncompressed Quicktime file, you could deliver the film DVCProHD, you could deliver the film in still images and they wouldn't know the difference.

I suggest that people continue to shoot in 24/25f for a film transfer.

That's exactly where I'm coming from, yet that is exactly the reason cited by DVFilm for their preference of 60i origination material. Agreed it is a limited scope of people who this would ACTUALLY effect in the long run, however it just would be nice to have which transfer house DOES transfer 24F to film on the searchable record, as it's more or less not, in case the situation arises for some LUCKY DOG here.

That would be my feeling too - except for the fact that it's Marcus, someone whose opinion I've generally come to respect, and was one of the first guys transfering the DVX's 24Pa to film, while some dragged their feet still asking for 60i. I would hope he has a good reason for requesting 60i, one that has to do with HELPING the customer - as opposed to conforming to their existing gear or selling software; the likelihood that it would be that or something else, like a predisposed bias against Canon, just seems unlikely to me. I wonder if any other film transfer house has looked at this 24F thing and come to a different conclusion.

Here's the thing - in rereading it, I'm sort of unclear as to what he is saying. It should be made clear as to whether he is point blank saying that 24F is simply unusable for film transfers, or just that 60i might do it (a little) better. That's a considerable distance between the two statements, as Chris pointed out, with 24F, the resolution is STILL better than 720P, which last time I checked that was the Varicam's resolution, and probably the most often used HVX resolution. Is it just that he's trying to help the customer the MOST out of the camera, but 24F "will do", or like my original question asks "Is 24F Not Good Enough For A Film Transfer".

Well, I would like to believe that too, but the facts are right there. There's is no technical reason why 24f is an unacceptable format.

None.

Even with reduced resolution I'd rather have 24 progressive frames than trying to extrapolate them from full res 60i. So should people throw out there 720p HD100's to shoot with 1080i cameras? No.

On a timeline in an NLE the DVX 24pa is no different than Canon 24f. So what would be the problem?

Reduced resolution? That doesn't hold water.

There is no good reason for requesting 60i from a Canon that shoots 24f. I can understand requesting 60i from Sony cameras, but not from the Canon's.

The people at dvfilm should know and understand the types of cameras and formats people will be using in regards to his services.

It is confounding. Probably should email Marcus so he can chime in as to his rationale for the why - especially considering that the resolution is higher, even with 24F on, than a HVX or Varicam in most circumstances. It would be interesting to find out what the others are saying. If these cameras are being treated like bastard stepchildren, like the Sonys rightfully deserves, there should be a clear why.

Excuse me, but there's no technical reason why 24f is an unacceptable format that we're aware of. There may very well be technical reasons that we are not aware of.

But no matter, because this is the point where a thread gets locked.

I really don't appreciate a technical discussion spilling over into accusations of a personal nature into the ad hominem realm of "ignorance" and "disservice" and "lack of professionalism." Such highly opinionated finger-pointing is simply NOT cool here, and will not be tolerated. I find it thoroughly disgusting in fact.

Marcus van Bavel is a valued, contributing member of DV Info Net, and I would never propose to tell him how to run his business, let alone know the first thing about the technical ins and outs of what he does for a living. Nor will I allow him to be attacked in this way on my forum.

If Marcus cares to respond and submit any input of a technical nature about the topic at hand, I will gladly open it up again and allow his reply... or Andy Young of DuArt for that matter, or anybody else who actually knows what they're talking about with regard to the fine art and science of HD to 35mm film transfer.

Until then, this subject is officially cooked to a crisp.

My apologies to anyone who was misled by that webpage. We certainly did not mean to say that 24F was not good enough for a transfer to film.

The comments (which I've removed now) were based on the assumption that 24F works the same way as "frame movie mode" on the canon XL-1, where lines on the interlaced sensor are blended together to make a non-interlaced image. For the XL-1 frame movie mode, we did do tests many years ago that show it's softer than 60i. We didn't do that test for the XL-H1, however, and we should have done so before updating that web page, and I'm sorry about that.

But regardless, the XL-H1 24F is definitely good for a transfer to film at DVFilm or anywhere else.

THANKS SO MUCH FOR CLEARING THAT UP MARCUS!!! Glad that it was simply an oversight - happens to the best of us. You know, you're a little like EF Hutton - when Marcus talks, people listen.

thanks for the response, marcus. i just came across this thread this morning, just in time to get an answer to this very question. timely! i'm sure that you're as crazy-busy as the rest of us, but it would be awesome if you could update your website to include some info on the Canon XH series of cameras. this is great info, and i'm sure you're going to get a lot of questions along this line.

Yes, thanks so much for your reply Marcus. I guess some here will sleep better tonight.

However, don't shoot the messenger...

24P or 24F was never really a technical issue as much as it was that Panasonic was willing to license the patent from Film Look and Canon and Sony weren't. As far as the technology itself goes there really no decernable difference.

It is not a simple process doing a film out from an F900 let alone an A1 or HVX200. No matter who says one format is as good as another, if you complete your project and then take it to be filmed out you are probably in for a big dissapointment. Other than the fact that you get to see some really big pixels...

If you plan on doing a film out go to the company of choice and pay to do a test before you start your project. Most of the reputable facilities have people who work closely with independent producers to make sure the process is well understood and goes smoothly. It helps them set your expectations correctly and it helps you understand what you need to hand them so that everyone has a pleasant experience.

The relationship that you build with this company is as important as the relationship you have with your DP.

Yes, perhaps a descending in ideal preference list for the Canon series, like the HXV's list. Something like....

1) HD-SDI output - either 60i or 24F
2) 60i to HDV
3) 24F to HDV (loses approximately 10-15% vertical resolution compared to 60i)

Well said. Thus ends another episode of "Resolution Blvd."....

Just to repeat a perspective that a number of us have mentioned plenty of times in the past but that doesn't seem to be sticking: when measured on static rez charts, 24F has a little less vertical resolution than the same camera's 60i. HOWEVER, when comparing 24F to any other 24-anything in the price class, 24F has at least as high, if not HIGHER resolution. No way, no how are you going to deinterlace XL or XH 60i material to 24fps and get better results than 24F. If you want better 24fps output, spend a LOT more money. Things change with time, but that's where the sub-$10K market is at present.

EDIT: In regard to Chuck's comments about the film out...I know nothing about the film-out process but I certainly wouldn't be surprised if less than optimal circumstances could lead to, well, less than optimal results. But done right, sub-$10K cameras can look great transferred to film. I saw Andy Young's film frames from a JVC 720p film-out at NAB last year and they looked beautiful. No reason that expertly done 1080p24 shouldn't look beautiful, too, especially if the source was well-shot 24F.

I would try DVFilm Maker. For example we show here that 60i can
be converted to 24P without reducing vertical resolution in areas
of the screen that are static, unlike with Vegas and many other
built-in deinterlacers:

http://dvfilm.com/maker/MakerVegas.htm

Thanks for the clarification Marcus. I may have overreacted in my previous comments and have subsequently edited them. I read a lot of assumptions about the 24f mode not only on these boards, but various others. It's a bit infuriating how little people actually know about the format and then try to pass off that knowledge as fact.

My apologies.

I've done video-to-film transfers at DVFilm with excellent results. The 35mm print looked great, projected at a huge drive-in theater screen. The original footage was shot with a DSR500 and a DSR250. I sent DVFilm a hard drive with the Avid QT. I'm confident if I shot the same stuff with the XH A1 in 24F mode, HDV, edited in a 24p timeline and sent that it, the transfer would be even better. The 24 fps progressive footage from the XH A1 is really nice, and actually looks better than very similar things I've shot with the same lighting setups with the DSR500.

Apparently the author of that info for the original post on here really didn't use 24f footage but made the statement based on Sony's version. I've see a lot of disinformation and just plain wrong information out there about this subject.

I was actually questioning the information that was on the DVFilm site, which lumped 24F in with Sony's 24P, as an inferior 24P process, until this very morning. I was wondering why that would be and posted it, thinking that if this were the case, it would be fairly common public knowledge. Turns out it wasn't, due to the fact that the original information on the DVFilm site was incorrect, due to an oversight. Marcus has corrected that page now, removed the 24F references, and I guessing will probably have a different page with updated 24F information. Upshot - 24F IS acceptable.

Cool. I was referring to the DVFilm info, not your personal comment; thanks for making them aware of things and posting their correction. They're good guys at DVFilm, and I assumed that whoever made the web post on their site had just used the Sony CF thing rather than looking at Canon footage, and apparently that's what they did. There's lots of that kind of thing that happens on the web, like...hmmm, it's got an F like CF, so it must be the same thing...or however the thought processes may work with some people.

Just because the F seems to confuse so many people, I'm gonna start referring to the XH A1's footage as 24P, which it is. I may say 24F when referring to the camera setting, but 24P when referring to the footage.

Sounds like a good idea. Perhaps that's what Canon's gonna do to, what with their having licenced "24P" and all, but that's a different topic altogether I guess.

Now here's the real topic that's been burning me up inside....

Kansas City BBQ vs. Texas BBQ!!!

Truth be told, I'll actually take Kansas City any day....guess that makes a traitor.

Well, I'll take KC BBQ, in the form of Arthur Bryant's specifically.

However, I also will note that Arthur Bryant was born in east Texas and when he died his body was shipped back there for burial. And...I am also a Texas expatriate and consumed vast volumes of east and south Texas BBQ in my childhood. So, when I say I like KC BBQ better, I guess I could also be saying that I like it because it's the way BBQ used to be in Texas.

How's that for a good weasel-worded answer. I might run for president.

Here's a platform for you - High Definition in EVERY home.

I'm glad this thread didn't get closed, and Marcus from DVFilm chimed in and corrected the error and cleared up the understanding.
Now this thread (as sooo many threads here) has valuable information again and a solution to someone's question :-)

We have still not answered the question.

We only know that DVFilm Maker does a better job of 24P conversion than Vegas et al.

How does 60i-24P compare with straight 24F in terms of quality?

It is not just the resolution that is the important factor. I'd be very interested in seeing a side by side of 60i-24P compared to 24F. More specifically I'd like to know how 50i-25P compares to 25F. Marcus any comments?

TT

Actually, my original question, "is 24F good enough for film transfer", has been answered. It does pose some other issues, and I'm just guessing with the volume of A1 and possibly G1/H1 material increasing, that would be reflected in the material/questions DVFilm receives, likely prompting a page with a recommended list of origination preferences - perhaps replete with screen shots as well. Perhaps a test even comparing the HD-SDI out / analog component out / 60i to HDV / 24P to HDV. But, the original question, the reason SO many people read this thread in so little time, was answered, and promptly at that.

My apologies. I missed Marcus' post confirming 24F is OK.

Still, I'd be interested in quality comparisons between 60i-24P vs 24F if anyone had performed such a test.

TT

Was made aware of this here....

http://www.dvinfo.net/conf/showpost....0&postcount=39

"60 i vs. 24F

XH-A1 60i:
824.3 Lines Horizontal (MTF50); 810 Lines Horizontal = Limit@Nyquist
674.8 Lines Vertical (MTF50)

XH-A1 24F:
810 Lines Horizontal (Limit@Nyquist)
590.4 Lines Vertical (MTF50)

Vertical losses: 12.5%
and less than 2% on horizontal resolution loss if beyond the 1440 restriction, slightly lower than the 60i."


While this quantifies the origination results, it doesn't really speak much to the far more important, and more subjective, issue - how does it look, and what's going to "look" the best; taking into account that some of us will place more importance on having 24 discrete frames at the point of origination at the "sacrifice" of the upper most resolution the camera has to offer - and gladly do so. That's the decision Barry Green said he would most likely choose, and I imagine most users would likely follow suit rather than convert from 60i, save the few resolution hounds. So I guess the question then becomes HDV vs. another way out of the camera during origination, say either the HD-SDI of the G1/H1 or perhaps even the analog component out of the A1, at 4:2:2, to an as yet uninvented Cineporter like device that accepts analog component in.

Todd

I am well aware of those resolution measurements but that is not what I am asking for. Perhaps I wasn't clear enough.

I am interested in the conversion from 60i to 24P and how that compares to straight 24F. I.e what is the optimal solution for someone shooting with Xl-H1/Xh-A1/G1 and transferring to film. How much resolution is lost in the conversion process? Is it worth the extra steps in the workflow??

TT

The measurments are useful in the secondary quantification you're describing. The 60i to 24P conversion would need to lose less resolution than the amount that 24F loses in order to be useful at all, and that doesn't really take into account the subjective issue of preference of 24 discrete frames from origination vs. converting 60i to 24P in the first place. I guess only Marcus, and your eyes, could answer whether, resolution speaking only, if his software would cost less of a resolution hit than Canon's 24F or not - if it would actually be worth shooting 60i. It would be an interesting shootout to see - exactly how to get the best image to film from the H1/G1/A1 series cameras.

i thought pete bauer already answered this question. more to the point, i think it is implied. 24F is 24P, and all we're really comparing is the quality of film-out among flavors of cameras, which i don't believe has been done. i get the impression that there haven't been that many film-out tests conducted using 24F yet, and especially not as direct comparisons with other cameras. that's a pretty expensive and time-consuming test. most people spending the time and money testing for a film-out are trying to answer the question, "how is my existing choice of camera going to perform under the specific shooting environment(s) of my particular film?" rather than generalized testing or camera comparisons.

The original question has been resolved, the 24F to Film issue - it IS good enough, which isn't what the DVFilm site said up until yesterday morning. It think the question now is the "Canon cameras versus themselves" - since it was brought up as sort of a side topic, how does one get the BEST out of the camera for a film transfer? Perhaps that should be new topic altogether, since the original question has been answered by Marcus.

I was NOT suggesting 24F was not 24P. I was using it to differentiate between 60i that had been converted to 24P by a software deinterlacer such as DVFilm and 24F(P) as recorded by a Canon camera.

I would have thought this would be interesting and useful information for someone delivering on film but shooting with a Canon. Perhaps I'm wrong.

I didn't think what I wrote was too difficult to understand?? Am I wrong?

TT

You're not wrong, and I'm pretty sure I get the crux of your question (see my post a few posts above) - I'm just not sure, even if the resolution loss going from 60i to 24P (say via DVFilmmaker) WAS smaller than the internal resolution loss of Canon's 24F, I'm not too sure that that would motivate too many users to forgo a 24P workflow, or origination. Since it has been resolved that 24F is acceptable for film transfer, I don't think anyone will look back on that issue. Now however, if it were proven that by going out the component output, using some as yet unnamed non-existent portable device, produced better images to film, as it might seeing as it is 4:2:2 full raster (according to Elton aka Barlow Elton), that might have a larger impact on A1 to film transfers at the very least, not as much so for G1/H1 since they have the HD-SDI.