I've been looking at threads posted on Panasonic's DVX 100, comparing it to Canon's XL1s. It seems that the DVX 100 beats the XL1s hands down with achieving a film look, delivering higher resolution (over 500 lines) picture quality etc. with the only draw backs being the fact that it does not have the option of interchangeable lenses and of course does not have the charming looks that the XL1s possess!

In short the DVX 100 will not allow you to modify your camera to comply to your specific production needs, e.g. focus pulling with a manual lens which is rather important for some filmmakers who need 100% control. The XL1s on the other hand does give you the freedom to modify your camera to suit your needs, which is one of the features that I love about it, but then it does not have the 24p frame rate, film look and high resolution of the DVX 100.

I’ve started my production company with a XL1s just over a year ago and I honestly love the camera to bits. I’ve come to a point now where I need to buy two more cameras for next year (before April 2004) and whereas my choice was simple, it seem that I’m not so sure any more. My production company no longer produces just wildlife television documentaries, we have just signed a deal to produce small television dramas towards the end of next year, and obviously here a film look would really be a good element to have.

Now what really turns me off from just going ahead and buying the DVX 100 is the lack of control over the lens when it comes to manual focus for focus pulling etc. But obviously the other features make it very seductive.

What I am really hoping for is that Canon will release a new XL1 series, which will do what the DVX 100 does, and more. I know there has been lots of talk, rumours and of course a wish list that went around, what I really want to know is weather anyone knows when this new camera is going to be released and of course what some of the features will be. Maybe it’s wishful thinking, but I hope this will happen before April next year. (It would really suck if I buy two new cameras only to have the new Canon released a month or two later!)

Any information (or gossip) would be greatly appreciated!

Didi,

Do a search on XL2. DVi Community even hosted a wishlist thread that was compiled and delivered to folks at Canon by our gracious host Chris Hurd.

The long and the short of it is: expect an XL2, but probably not until mid or late 2004. A sign that the XL2 is just a month or so from being announced will be an MSRP drop on the XL1s.

Thank you Robert.... will go and have a look!

The wish list idea is great, but I wouldn't pin too many hopes on it. Witness the recent Sony announcement of the PD-170. They were obviously looking at a completely different list and only included minor tweaks to the existing PD-150. This seems to have caught everyone on the various forums by surprise.

I see that the DVX offers a MANUAL PULL FOCUS option? As well as ana and widescreen adapters?

The Problem with the manual lens focus pull is that the focus ring does not have any stops on it, very much the same as the standard auto lens that the XL1s has, you can turn it around till you're blue in the face which makes marking focus points a rather tedious exercise. Fortunately with the XL1s you have the option of the Manual lens which of course gives you full control, and that’s what I like about the XL1s.

There are four basic things about the DVX100 that I do not like and which makes me wanna stick to my XL1s, 1) The focus ring that does not have definite stops, (main reason) 2) The limitation of adding "professional" accessories like various lenses. (Important, but not really that important) 3) The plastic fragile feel of the camera and it's looks. The XL1s at least looks more professional and feels like it can handle the knocks and bumps associated with normal production work. (Obviously just being picky!) 4) The stabilizer not being as stable as the XL1s's, face it, Canon's stabilizer on their 16 X IS II is as smooth as butter and I think it will be hard for any camera to beat!!!! (Something that I would surely miss!)

But then there are three things which I like about the DVX100 and which I really hope the new XL2 camera will have. 1) P24 frame rate with the "real film look" that everyone’s raving about! 2) The quality of the lens (glass) which delivers much higher resolution than both Canon's Manual and Auto lens, meaning you don’t have to buy expensive third party lenses and accessories to get the same quality on the XL1s that you'll get on the DVX100 as a standard feature. 3) No quality loss when shooting in frame mode, it would be nice to keep the full quality that your lens and CCD chips can deliver on the XL1s which seems to be the case with the DVX100.

That is basically the only three things that I'm hoping for, of course all the other features that where listed on the "wish list" like the programmable flash card with tweaking camera settings, true 16:9 CCD chip capability and the XLR connectors with phantom power will all be more than welcome additions!

In short, give us a camera with all the handy features the XL1s comes with, and delivers the same if not better quality which the DVX100 delivers, that goes for the amazing film look and picture resolution.

I suppose the BIG question is when... Some say August 2004, which just makes me wanna cry, other speculate early next year... and I'm of course holding thumbs for that!!!! Only time will tell...

I think the earliest you might get an announcement is at PMA or NAB. Both are in the Spring. The earliest the camera would ship would be summer of 2004. These are just guesses based on past Canon new product releases. A rebate on the camera would also predict a new model, as Robert said.

Didi I'm in the same boat and I'm sure many others are as well.

I was dead set on buying the XL1s but I have reversed myself and am about 80% sure I'm gonna go with the DVX. As much as I love the way the XL1s looks feels and acts I can't wait much longer...Out of box the DVX has xlr connectors, stock wide angle lens, cinema gamma and full rez on movie mode... As much as I hate to say this but for my dollar the aging xl1s can't compete.

<<<-- Originally posted by Didi Schoeman : 1) The focus ring that does not have definite stops, (main reason)-->>>

http://www.centuryoptics.com/products/prodv/dvx100/Manual_Focus/manual_focus.htm

Century Precision Optics has introduced a manual focus ring attachment that provides hard stops for the DVX100's focus ring, as well as industry-standard pitch gear teeth for use with follow-focus and jib head focus motors.

Thanks for that info Barry... it sure made me sit up and listen and I'm looking at the camera with different eyes!

Fortunately I still have a little time till April... not a lot but a lot can happen in 6 months. If the next XL is not on the market by then, (and of course able to compete) my bet might just be on the DMX.

This is not gonna happen, but....

I figure since Canon doesn't have a professional line to kill with any advanced features, they are the ones that could get away with this....

In addition to all the wishes everyone has noted, the XL2 should simply take 2 more color samples 4:2:2 and have a setting to ~allow~ streaming uncompressed 4:2:2 out the firewire(s) port.

I understand that this is not the DV standard and it would need a special driver/codec or program on the machine side, which would be provided by canon, for capture. All the XL2 has to do is throw away the two extra color samples to get 4:1:1, compress 5:1 with DV as usual and write it to tape. The tape would always be DV of course, but the camera would be able to output better SD video through firewire, which certainly has enough bandwidth for this.

Even if they don't want to do 4:2:2, just the uncompressed 4:1:1 output is worth it! If they keep the XL1 design, the professional setup certainly allows for mounting a harddirve on the camera for high quality, no compression capturing, or you could hook it up to a digiBeta deck when in a studio.

They do this, and every other camcorder is history, at least for indie-film production.

Juan

Sharp thinking, Juan. XL2 + laptop = prosumer version of the Viper HD system, and Canon wouldn't even have to engineer an onboard hard disk recording system.

Yeah, that would be fantastic. We were discussing this in another forum and the only reason why we thought companies wouldn't do it is to avoid out-staging their more expensive pro products. Canon doesn't have this problem.

I'm also an electrical engineer, and I can't see the huge obstacle in doing this. The CCD's are there, all you have to do is take a couple of more samples(perhaps faster clocking), have some extra memory to buffer them, and come up with a simple scheme to transfer them via firewire, if there already isn't one.

Canon can even include their own PC/Mac software to record video, and then you can use the output footage, or whatever. Of course, program manufacturers are sure to add support for it quickly.

Like I said, it's probably not going to happen, but I can't help saying it...it seems like so little technical effort/cost to blast every other manufacturer away. Imagine...4:2:2 uncompressed video! Even if the XL2 turns out to be SD it would be awesome.

Juan

The idea sounds awesome Juan, I'm just curious about the fire wire cable and what the maximum length of it is before it becomes ineffective? I'm not an expert but I'm sure I've read somewhere that 4 meters is the maximum length of a fire wire cable... which would be a pity really, especially in a studio set-up...

I am also curious about your viewpoint on the current Canon lenses with the setup you described, I suppose an upgrade in glass quality would be necessary to up the overall resolution of the lenses to match the output of the CCD chips (to at least 500 lines), or alternatively I suppose one could use third party lenses, but then it becomes expensive…

Yes, the firewire cable length is limited just as always...however, there are many ways to get around this. You could have a laptop to record near the camera, or you could use on of those on-camera harddisks for recording. Given the right output you could also hook it up to a D5 or beta deck and get better quality than the miniDV tape.

Since what I am suggesting is to output uncompressed 4:2:2 SD video, i am not sure if the lenses the XL1 has now would be any trouble. The advantage of this idea is that the output will be uncompressed, with no artifacts, and it will have better color due to the two additional samples. True, the resolution and general image is limited by the glass, but even with the standard XL1 lens that everyone has, the uncomp. 4:2:2 output will look a LOT better, and will make much better use of any lens upgrades.

Think about it...the data IS there..the only reason why you loose so much data is because of the DV 5:1 4:1:1 format. The idea doesn't require any changes to the lens/ccd hardware, just the software/hardware that samples the CCD voltages, and manages the firewire port, along with the standard modifications to the menus to add this option, etc.

I read somewhere that the DV scheme throws away 80% of the data that is collected from the CCD's. I think that if we pay for the CCD's and lenses, we should be able to get all that we can out of it somehow. I know why sony and other companies won't do this, but Canon has a good oportunity to do it now with the XL2.

Juan

I've just learned something new! It would be awesome if they could do that... but I suppose it's one of those wish list things which will remain on the wish list... pitty really...

are for sale on e-bay in contemplation of an XL2....

Jeff, it would seem that they are going for a song! Found a place which are selling brand new XL1s PAL cameras for $3099.00 (NTSC models are about $300.00 more)

Things are looking rosy!

Who knows we might just have Christmas early this year…

There are always a lot of XL1(s) for sale on ebay. Nothing unusual in what I see.

The biggest problem that Canon faces, is the same as the others: The lack of uniform standards between them and the other manufacturers. Every body wants their standards excepted as THE one and only.

It wouldn't be wise in this world wide down business climate, to retool a plant, and, then, find that different standards are now in vogue by all the others.

I believe that the current meetings among several of the manufacturers will eventually solve the problem. How much lead time do they need to arrive at THE standards, I do not know. Six months? A year? Two?

I am not holding my breath on a new XL-2s. An upgraded XL-1s………, maybe.

Sony's memory stick appeals to me more than tape or disc. I love the idea of no transport system.

No moving parts, no break downs.

Here's another one of those ideas for the XL2 wish list, which i've never seen mentioned anywhere.

The idea is for a means to simulate the latitude of film. We all can identify DV when we see those washed out whites, because of the linearity of the CCD and the fact that it can't display such a wide range of brightness levels as film on a single exposure..

I think i've figured a way to get the same latitude of film and get the same look...two ways of doing it.

The expensive way, is to have an additional CCD, set to collect less light than the main CCD by a certain amount. An algorithm would control exposure and maybe an auxiliary iris on the secondary CCD. By means of software, a processor could combine the darker image from the secondary CCD(thus better capturing highlights) into the washed out sections of the primary CCD image.

The cheap way, is to use the same CCD but instantaneously capture a second image after the primary frame is captured, but with less exposure. Not sure if this would work, due to the fact that for a certain shutter speed, a shorter exposure of the same image would follow to get the detail in the brighter areas of the scene. I guess you could do it with the iris as well, but this would change dof....the change in iris opening would be minimal so this might not be a problem.

It would be an interesting experiment, i think i'm going to try it with one frame and see how it works....i can compare it with an image of the same scene from my 35mm still camera.

Juan

Digital blending of frames is a well known technique in still photography. You can find a tutorial here (http://www.luminous-landscape.com/tutorials/blended_exposures.shtml). The only problem is the scene can't contain any motion

As Jeff points out, the problem with the two-exposure method for achieving high dynamic range in video applications is it doesn't work for moving subjects. (Although, you can use this method for achieving high dynamic range stop motion movies of very slowly moving subjects, such as clouds migrating across a daytime sky. I worked on such an HDR time lapse camera as part of my master's.)

While the contrast compression method described by Jeff's link will produce some pretty pictures, it's not photographically correct, because it doesn't take into account the gamma curve(s) of the camera, and moreover, it won't increase the actual dynamic range of the image, since the same number of bits will still be used to describe the radiance of each pixel. See Recovering High Dynamic Range Radiance Maps from Photographs ("The HDR Paper") at www.debevec.org (http://www.debevec.org).

Thanks for the input...you guys know way too much :)

I'm not sure exactly why it wouldn't work with 2 CCD's? Like if you split the light beam from the lens 50x50 and then had different exposure settings on both CCDs, on of them more sensitive to capture the darks and the other less sensitive to capture the brights that would otherwise be washed out.

Why wouldn't this work for motion?

Juan, this would work, and those cameras that do motion HDR use just such an apparatus, with a few caveats. When I was at the ICT Graphics Lab (http://www.ict.usc.edu/graphics/) in Summer 2001, producer Jamie Waese nailed two VX2000s to a 2 by 4, attached a beam splitter with ND filters of equally spaced stops affixed to the facets to the lens of one of them, and created a real-time HDR lightprobe (http://www.ict.usc.edu/graphics/reserach/rtlp/index.html). (A patent was filed for this; I don't know whether it was ever issued.) The setup was not without its issues, however, one of them being a slight chroma shift on some of the beam facets, a problem which had to be corrected in software, the overall effect being a slight degradation of the resultant HDR video.

Since Canon and SONY are already experts at split-beam imaging, having produced a number of 3 CCD video cameras, I'm sure with the proper engineering either manufacturer could produce a camera capable of higher dynamic range without exceedingly prohibitive additional cost. But, since there is a market for this (it's called HD cinema), one might speculate that the reason why SONY hasn't already implemented such a scheme in its high definition line of cameras is that a more effective and economical way of achieving higher dynamic range is to use better CCDs, a luxury prosumer equipment doesn't offer.

"Yes, the firewire cable length is limited just as always...however, there are many ways to get around this. You could have a laptop to record near the camera..."

Just so you know, they make firewire cables with built in repeaters that go pretty far. We use two that are 150ft. each and are looking at getting another that's 200-250ft long. They're of course more expensive, but thought I'd let you know they exist in case you didn't.

Thanks Robert! Well, i guess if Sony patented the idea then Canon might have trouble applying anything similar to it in their next camera.

I'm curious as to why canon has 'held back' as far as video quality goes on the XL1 series. They seem to have plenty of resources and experience, and they don't have a professional line to mantain, so they could go all out on their next XL upgrade, adding features that until now Sony/Panasonic/JVC wouldn't include in order to not overlap with their pro line.

I don't know, maybe they are selling so many XL1's just based on it's unique design that they are not really concerned with trying to sell any more. After all, the XL1 is in a class all by itself anyway. It just seems really strange that the GL2, which is supposed to be a step down from the XL1 has more pixels per CCD's than the more expensive big brother. I really hope they don't disappoint with whatever the XL1 replacement will be.

Juan

Just for the record, the ICT's Hight Dynamic Range Light Probe was not a SONY endeavor.

Back on the topic of uncompressed output on DV camera's... I know of people that hacked their Sony DSR-300 DV camera to output UNCOMPRESSED video through one of the analog video outputs...

Aparently it takes a bit of soldering to activate the analog video output during recording video, since by default it's turned off while shooting, and then turned on on playback from tape, with a compressed signal.

They'd hook it up to a G4 with a RAID system, and captured uncompressed video from a DSR 300...

It wouldn't be suitable for everyone, but I can imagine I'd bring a system on-set if it would allow for uncompressed video output from a DV cam. It would be even better if it would have firewire or SDI output of uncompressed video :-) Buy a Decklink board and you're set.

Bar3nd

That's very interesting! I've been wondering if I somehow obtained schematics or had an 'extra' DVX to open up carefully and see if it could be hacked in a similar way. I guess the thing to know is if the CCD chipset does DV compression internally right after sampling, or if a separate set of chips does the sampling/compression. Perhaps, even if it is done on a single chip it can be switched modes somehow...i wouldn't mind inacapcitating the tape if i can get uncompressed video out in a studio setting....

I'm curious as to why canon has 'held back' as far as video quality goes on the XL1 series. They seem to have plenty of resources and experience, and they don't have a professional line to maintain, so they could go all out on their next XL upgrade, adding features that until now Sony/Panasonic/JVC wouldn't include in order to not overlap with their pro line.

Canon has a lot to lose. Canon and Fuji are the only manufactures of high end broadcast lenses. Fuji does not have a video camera presence. If Canon were to give away high end features in low end consumer market cameras, it probably would impact their much greater profit BQ lenses.

Ah I see. So the high end broadcast lenses work on the XL1?But If not why would it impinge on their broadcast lens market? I mean after all, even if they make an XL2 that can output uncompressed HD video, the factory lenses will be nowhere near their broadcast quality lenses....or do they make broadcast cameras as well? How does the DV camera capabilities interefere with their lens market? You'd still have to buy their high end lenses to get good glass....

Juan

Juan,

<< So the high end broadcast lenses work on the XL1? >>

No, they do not.

<< But If not why would it impinge on their broadcast lens market? >>

If Canon were to give away high end features in lower end consumer market cameras such as the XL1, this would take sales away from professional Sony, Panasonic and JVC cameras, which all use Canon broadcast lenses. That's how it would hurt Canon's broadcast lens market.

<< I mean after all, even if they make an XL2 that can output uncompressed HD video, >>

The forthcoming HDV format which undoubtedly will include the XL2 will *not* be uncompressed. It most definitely will be compressed.

<< or do they make broadcast cameras as well? >>

No, they do not.

<< How does the DV camera capabilities interefere with their lens market? >>

Explained in first part of my response, just above. Hope this helps,

> Canon has a lot to lose. Canon and Fuji are the only manufactures of
> high end broadcast lenses. Fuji does not have a video camera
> presence. If Canon were to give away high end features in low
> end consumer market cameras, it probably would impact their
> much greater profit BQ lenses.

Even more to lose: imagine what would happen if that other manufacterer were to stop selling CCD chips to Canon.

Chris,

When i said uncompressed I meant an option to output uncompressed 4:2:2 or 4:4:4 via firewire, even if the on-board format is less than this.

I'm in the process of doing an experiment with a TRV900, and all the data remains 4:4:4 way until it is DV encoded for tape and firewire output. I have all schematics for this camera, and I don't know for sure, but I would bet that the XL1 and just about every other camcorder works at full quality all the way until compression/decimation.

I am planning to do some data capture tests by serializing Y/C 4:4:4 uncompressed data after it is adjusted(but before it is decimated/compressed of course).

Right now i'm working on the circuitry for recording the data, i'm using 2 IEEE 1394 interfaces and high-speed shift registers to serialize the data to hard disks.

Anyway, my point is, the data is there, despite what limited tape format the camera uses. If they include a feature to somehow access the uncompressed/undecimated video, it will be an incredible feature. After all, in this age of computers why should we be limited to the tape format's capabilites.

Juan

After all, in this age of computers why should we be limited to the tape format's capabilites.

The manufactures don't want to give away the high end features and corresponding benefits. The companies have an obligation to their stockholders to produce a profit. Profits will diminish if pro features are incorporated (or enabled) on entry level products. It may seem unjust or unfair, but basically it's all about features, benefits and making a profit. After all, if there is no profit, there's no R & D to develop and refine technological advances.

"Dumbing down" the features on the upscale software is standard practice in NLE's. Avid and Cinewave give you "Upgrades" by unlocking the software potentials you already own.

Yup, it's about profits, and marketing. That's the biz

That is all true. But the only camera i have seen this feature on is the Thomson Viper. If it is a matter of outstaging professional products, why don't they include this feature in the Panasonic SDX900 for instance.

I beleive that there are PLENTY of other features that you just do not buy when you purchase a cheaper camera like the DVX100...large, better CCD's, and the glass. This will never change. But I think that not allowing you to use the the capabilities that are ALREADY THERE, is just something that whenever one company does it, all of them will.

Think about it, even the XL1, or the DVX100 will not have the larger CCD's, low light performance, more lines of resolution and the great lens of higher end cameras, even if you could get 4:4:4 uncompressed output out of it. Not to mention the ability to record on professional casettes, etc. This feature will only work tied down to a laptop or even using an expensive additional harddrive. Actually, even though you may get 1/3 CCD's, as far as color resolution you are getting the color from a CCD 1/4 that size because of the quantizing. You bought the thing, why shouldn't you be able to get the output from it? If all camcorders had this feature, they'd still be in the same range. A DVX100 wouldn't be any more better than an SDX900, if they both have the 4:4:4 uncompressed feature.

Anyway this is just my opinion, you guys are probably right. I just think it's like giving you a ferrari for less money but limiting the engine output with a set screw. All you have to do is go in and turn it.

Juan

btw. for XL2 they should make firewire 800mbps (100MB/second) instead of 400mbps (50MB/second). that'll take care o uncompressed real good =D.

<<<-- Originally posted by Juan P. Pertierra :Think about it, even the XL1, or the DVX100 will not have the larger CCD's, low light performance, more lines of resolution and the great lens of higher end cameras, >>>

Well, possibly the last unique feature of the XL1 in the current market is that it is the only camera in its class that allows for interchanging lenses. With the proper adaptors, the best lenses in the world can be utilized.

Good point! Just because you can change lenses on the XL1 just like the pro units, that doesn't mean everyone is going to buy the XL1 and not the pro ones.

There will always be features that cost money, and there is no way around them but to buy a more expensive model, that costs more to build. Piping 4:4:4 video out doens't cost them any more other than a simple design change and maybe some simple software.

If you were to use an MJPEG codec or something with a datarate
between 10 - 20 MB/s that would be doable to store on current
harddisks and increase colorspace dramatically... Uncompressed
SD is only 30 MB/s.

Canon, Sony & Sharp Corporation have just signed a standards agreement for domestic Hi Def. Anyone know anymore? Cheers from Oz, Pete Constable

Well, you've hit it right on the button--there's a possibility the XL2 will use the HDV standard--which would be unfortunate, because Canon could surely come up with a better format free from the constraints of the MiniDV carriage.

Pete,

Where did you hear this? Was there a press release like when they agreed on the HDV def? Would love to see the link to this news...hopefully we are closer to the HDV world than we think!

Clay

Clayton, I think that's what Pete's referring to. There's even a domain dedicated solely to the September 30th announcement. http://www.hdv-info.org/

The official site is

www.hdv-info.org

but there are no details posted yet, just the press release.

Canon's involvement in this, together with the advent of lower cost CMOS sensors and the fact that the XL2 is not predicted to come out for at least a year or so, i'd have to say that this all points to it being HD.

I think an impressive new product that shows more or less what direction this all is going is the Kodak Pro14n(?) digital SLR camera. It has a CMOS sensor the size of a 35mm frame and captures something like 13megapixels, and costs somewhere around $4k.

<begin joke>
now, the burst rate is at least 3fps, so we should all chip in and get 8 of these, hookem together in front of a good lens and a mega beam splitter. voila.
<end joke>

Seriously, it would be nice if it had at least an analog video out but that created NTSC video using some of the chip resolution...but i think usually they just have a really low res(320x240?) output...might be wrong.

Juan

I think an impressive new product that shows more or less what direction this all is going is the Kodak Pro14n(?) digital SLR camera. It has a CMOS sensor the size of a 35mm frame and captures something like 13megapixels, and costs somewhere around $4k.

This camera has very few satisfied users. It's low light performance is dismal and suffers from a severe noise problems under some conditions. The Canon 1Ds is a much better camera by all accounts but is several thousand dollars higher and still only 3 fps.

Jeff,

What do you know about the Sigma SD9? Have heard some good things, but also have learned not to trust everything I hear! I know it only records in "Lossless compression RAW data (12bit)" where as the Canon 1D does RAW data 12bit and JPEG. Which would you recommend?

Clay...