Wide Shot Resolution
 

I have been recently shooting with the 3x Canon zoom. Great lense that I love for hand held shooting.

But when I shoot a wide vista I notice that image doesn't appear as sharp as when I shoot a near object with the same lens. Both are in focus but I think it may be that DV (or the Canon system) just doesn't resolve images where the distant objects get so small in frame such as in a wide vista.

Has anyone else noticed this?? Any solutions??

Sounds like a well-know limitation of DV. The heavy compression takes its toll on scenes with a lot of complex detail, like landscapes. Also, as things get smaller there are just not very many pixels available for them. Always a frustation of mine as well.

Last year I did a big project that included photorealistic computer rendered landscape shots as well as real footage. Even though there was no "lens" used in the computer sequences they exhibited this same problem (although the same frames rendered as JPEG's looked much better). If you had a chance to see "28 Days Later" in the theatre on a big screen you would also have noticed that the landscape and cityscape shots looked pretty ragged, although I thought many of the closeups and medium shots looked surprisingly good.

So I guess we just have to wait a little longer for the HDV market to heat up and offer us a few more options :-)

Cameras with smaller chips will be softer on wide shots, but the lens and back focus adjustment can also contribute to the problem. If the shot is so soft it looks out of focus, then it's probably a back focus issue. Professional cameras have lenses with back focus adjustment; others have to be sent in to a repair facility. However, what you're seeing is most likely just the way it is with smaller chips. On occasion I've done two camera shoots where the second camera is a smaller chip one than the main camera. In those instances I do the wide shots with the bigger camera and closeups with the smaller one, and they usually match fairly well.

That's what I figured. I'm working with a director who has never shot on DV and is voicing concern about the wide shots. I don't believe it's a back focus issue either but simply the DV format as suggested.

Has anyone tried different diffusion to help? I know this sounds odd but I'm from the film world and if we have a moire edge problem sometimes a little diffusion helps soften the edges so it is not as apparent.

You're talking about two different issues. Over enhancement is something else, and it can be turned up too high and get that "edginess" that some people associate with DV because they haven't seen DV shot with fully professional cameras that are properly adjusted. The soft wide shot issue, however, is a function of the lens and camera you're using. On a DSR500/570, the wide shots look pretty good. On a 1/3" chip camera, they don't.

So it's an issue of format resolution not a "video in general" issue.

I was hoping that I might get lucky with the diffussion route. When you transfer film to video, edges sometimes buzz or moire depending on the size of the lines in the frame. Shooting a closeup of the old Amex card was a classic example with the small lines on the card. Adding diffusion when shooting often helped. When the Spirit film-to-video transfer machine came along it solved most of those issues.

Thanks.

Bill, you make some excellent points, and thanks for challenging what I believed to be the prevailing "wisdom" on this topic. You got me interested enough to do a few quick experiments, and they would tend to support your statement about optics and chip size being more of a factor than the DV compression itself.

First I took a landscape shot from my Nikon 5700 (a 2/3" 5MP still camera), resized to 720x480, dropped it into FCP, rendered as a DV compressed Quicktime file and finally exported back to a 720x480 still. There actually was very little difference in the results, much to my surprise.

Wanting to quantify this a little better, I did the same thing with a JPEG of the EIA 1956 test chart. There are some relatively subtle differences here if you magnify the image, but not much. I then took a still frame which was exported from DV footage of a high-resolution print of the same chart as filmed with my VX-2000. Wow, the difference there is striking when you enlarge it. You can see for yourself here: http://www.greenmist.com/pdx10/chart/res.jpg. Now I don't claim this to be terribly scientific, but it's enough to convince me. These images are cropped from a small portion of the upper left hand corner of the chart and were originally ~74x128 pixels. I first set auto-contrast on them and then enlarged by 300% using Photoshop bicubic interpolation.

Alas, now I'm depressed! I had always thought the real culprit was the DV codec itself, but now I'm convinced it's the chips in our prosumer cameras...

" I had always thought the real culprit was the DV codec itself, but now I'm convinced it's the chips in our prosumer cameras..."

The DV codec can give far better resolution than we give it credit for.
I recently had some material shot on Digibeta transferred to DV for me to edit into a package. The quality was far far superior than the image obtained from my XM2.
Cheapo "prosumer" cameras really can't cut the ice - whether it's because of the CCD itself or the circuitry, firmware, call it what you will, but DV itself is capable of very good pictures.

Robin.

All you have to do is look at the differences between a 1/3" chip camera and a 2/3" chip camera, both DV, to see what it's all about. Interestingly enough, to my eye, anyway, there seems to be a bigger difference between a 1/3" chip camera and a 1/2" chipper than there is between the 1/2" and the 2/3" ones.

Interesting test.

Does the interlacing acquisition of video have anything to add to this equation? Is it a different method of acquisition in a digital still camera versus digital video?

As to the 1/3" to 1/2" being significantly better than the 1/2" to 2/3" chips, that makes sense to me because there is a great relative difference. In the first comparison you are starting at 1/3" in the second comparison you are starting at 1/2". If there were a 1/4" 3 chipper there would be even a greater difference compared to 1/3".

It also may have to do with a threshold of acceptance for professionial imaging. 1/3" is the lowest threshold IMHO but is becoming much more accepted lately. That doesn't mean that any camera in any format can't make interesting and beautiful pictures. Just that each format has it's pro's and con's as we witness in 1/3" DV wide landscapes images.

Well, a CCD chip has only so much pixels to see differences. The
wider you go the faster detail changes on a smaller scale
(usually). So I'd say at some point the CCD's just can't see the
extra information anymore.

As for the camera's lower quality compression. I'd say this is
mainly due to processing power available. The DV codec inside
such camera's is probably just of a lower quality than the DV
codecs we use in our PC's, simply due to processing power. I
have no proof for this, but it does make sense. First they need
to combine 3 datachannels into one which is already 30 MB/s.

Then they also need to do this quite resource heavy DCT
compression algorithm and lay it to tape all within a second for
30 frames. That's quite a few things to do. And I'm not even
counting things like white balancing or other effects etc.

So I would not be that surprised if they cut some corners on
implementing the DV codec.

This is all just a guessing on my part, ofcourse...

I guess that's why some people are looking to capture raw uncompressed video. See this little thread (800 something replies):

http://www.dvinfo.net/conf/showthrea...threadid=20332

DV encoding/decoding seldon shows it's shortcommings in terms of resolution..the high resolution parts of the image get a digital value (if above a "visible treshold") just like the "low resolution" area's. So resolution as such doesn't belong to the DV codec. Resolution is limited to 720x480 (576) points for the "ideal" camera. Rather artefacts like blocking, quilting, mousquito noise... are the DV shortcommings. There are differences in DV encoding quality. Once the DCT tranformation has taken place (a mathematical process equal for all cams) come the tricky steps like optimal quantization table choice, motion estimation...and these approaches ( not fully defined in the bluebook) can make the difference. Again,.. not in terms of image resolution

I have just bought a new Panasonic DVC30, and all this talk about chip sizes makes me depressed seen as though mine is just a ¼ incher :(

The 1/4" chip camera will look good on its own. People have done documentaries using 3-chip 1/4" chip cameras. It's only when you intercut the footage with bigger cameras that you start seeing a difference. Or, if you've been watching TV with high end footage and then see something shot with a smaller camera, you will think it's soft. Eventually, though, your eye gets accustomed to the look of whatever you're watching. Anyway, you don't shoot movies with video if high resolution is your main concern. That's what HD and film is for. Your 1/4" chip camera is probably about the same quality as the old VX1000, and look at all the world famous art house movies made with that camera.

<<<-- artefacts like blocking, quilting, mousquito noise... are the DV shortcommings. -->>>

And that's what I always had heard as well. But in my test these do not appear in the JPEG of the chart that was compressed with the DV codec however the do appear in footage of the chart that was filmed with my VX-2000.

Whatever the reason, these artifacts must be related to the camera's chips, hardware or firmware. But to put this all in perspective (trying hard to snap out of my depression here ;-), that was a tiny area of the full image which was enlarged 300%. The full chart from the VX-2000 looks like this. But what surprised me was how just how clean the DV-compressed JPEG looked. To me, the main difference between it and the original is the limited luma range which shows up in the darker gray background.

Just watching Eurochampionship football on TV.
Wideshots of the entire pitch: the players show up like blotches, definitely soft.
Close shots are brilliant in definition.
Same thing. Big expensive cameras. (New TV)

Boyd tell us somewhat more about those enlarged pics. I suppose the first one is an original (electronic) JPEG image of the chart. Size ? (KB). The second one, recompressed in the DV format? Full frame? The thirth one: a VX2K pic taken from the chart?.Looks to me as a line doubled field (low vertical resolution).

Interesting. I think that Rob is on to something, makes me wonder if (non-mechanical) image stabilization would be part of the cause of this degradation?

Andre: For the camera test I used the chart from John Beale's website which is 3458x2608 pixels to print 216 dpi hardcopy at 16"x12" on high quality inkjet paper with an Epson Stylus 1280. Shutter speed 1/60 sec, manual iris with zebra to determine exposure. Camera was ~10' from the chart. I shot that over a year ago however, I wish I was viewing the full frame at the time because it could have zoomed in slightly. I think the VX-2000 image is in the same ballpark as tests I've seen elsewhere:

http://www.pixelmonger.com/chrt_pd150.jpg
http://www.bealecorner.com/trv900/respat/trv-res.jpg

My JPEG on the left is the same file used to print the hardcopy, but resized in photoshop to 720x480. The middle one is that JPEG dropped into an FCP sequence, exported as a Quicktime DV-compressed file. I then opened the movie in the Quicktime Player, copied a frame and pasted into photoshop.

All of the images are 720x480 and I just enlarged and cropped a representative section to put online.

Richard: if it is a non-optical image stabilizer, then yes, it will
definitely affect quality and/or resolution.

Thanks Boyd for the additional info. Still two questions:
-Did you drop the VX2K picture (image sequence I suppose) also into FCP and further followed the same procedure as for the middle pic?
-Is QT outputting (deinterlaced) frames?
I am still puzzled about yr results: allmost no dif between a 3458x2608 JPEG and a 720x 480 DV image for the left two pics, and the large amount of mousquito noise and quilting on the right pic.

Andre: as I said, the VX-2000 clip is a little old, but I actually filmed to tape, captured in FCP then exported as a JPEG. Sorry, I don't remember further details. Maybe later today I'll capture another example from both my PDX-10 and VX-2000 under more controlled conditions and see if there's any difference. But like I said, try enlarging those charts in the other links I provided. I see the same sort of mosquito noise there as well.

And just to clarify, the hard copy of the chart was 3458x2608. The two JPEG images on the left were first resized to 720x480. But regardless of the quality in the VX-2000 example (as you note) the surprising thing is how clean the DV compressed version of the original JPEG is.

A new test
 

OK, I re-shot my test chart. I also tried the same test with my PDX-10 and included both results. The new test can be seen here.

This time I was careful to frame the shot as exactly as possible so the chart filled the whole frame. I was about 6' away and shot outside in the shade with indirect light on a cloudy day. The video was captured directly via firewire to my laptop running BTV Pro. I then dropped these clips in FCP and exported still images as JPEG's at the highest quality. I did not deinterlace.

I used manual white balance to a blank sheet of the same paper as the chart printout (which was again 16"x12" @ 216 dpi). I zoomed in to the max and manually focussed, and tried autofocus as well with no apparent difference. Then I zoomed back out to fill the frame with the chart. One interesting thing, the VX-2000 couldn't seem to white balance properly under these conditions, the result was very bluish. Both cameras were shooting interlaced at 1/60 sec shutter. The VX-2000 was around f6 using the ND1 filter. On the PDX-10 it doesn't tell you the exposure and it automatically drops in its undocumented ND filters. I converted the results to grayscale and used the Photoshop auto levels function. Once again I cropped the same area of the chart and enlarged 300% with Photoshop bicubic interpolation.

So anyway, these tests yielded better quality images and dont have the blocky artifacts of the old test (I'm guessing the problem with the old image was that JPEG quality was set low when I exported from FCP). However they don't really change the conclusion that the camera is the main limiting factor in image quality and not the DV codec.

Thanks Boyd for the interesting pics. Cams do indeed limit resolution. But that second (non cam) pics still puzzles me...Did you really drop the resized JPEG testpattern into the FCP timeline, made a DV rendered AVI file (to get the DV conversion) , and exported a frame out of this AVI file?

Yep, although it was Quicktime... isn't AVI some sort of PC thing? In fact, I tried the test two different ways since the first time it looked too good to be true. I was afraid that FCP was using some higher-quality internal representation rather than DV when I exported directly from the timeline. That's why I rendered and exported as a DV-compressed Quicktime movie, then used the Quicktime player (a separate application) to copy the frame and paste into Photoshop. I got similar results using this same methodology with a photograph of a landscape. And that's what got me started on this whole thing. I was pretty sure that if I took a clean JPEG of a landscape with lots of fine detail that it would fall apart under DV compression. Much to my surprise it did not. Guess I need to start saving for a DSR-570 ;-)

Try some tests for yourself and see if you get different results...

Shocking!
Where is all that noise coming from?
Where is all that resolution going?
Boyd, is it possible that the 'sharpness' setting on your VX2000 and PDX10 are set way down? Could it be that we are setting them too far down? Notice: I include myself. Also, is it possible that the AE point is also set too low? I am sure you probably set it down too, like I do, so the image might be somewhat underxposed to begin with, this might account for the noise. Also, do you have FCP set up do it's math on the maximum bits per channel?

Boyd I did some tests the in the past using the eia1956 and zoneplate testpatterns, maybe I'll have to do this again, but I never got a pic out of a DV stream that good. What is the filesize for yr DV converted JPEG pic? As far as I remember it should be less than 80KB.

Iganacio: I always keep my PDX-10 sharpness at the minium, not sure on the VX-2000 and it isn't handy right now, but I suspect it was a notch above the minimum. You are probably right about this influencing the results, hadn't thought of that. However most of the time my goal isn't to produce sharp resolution charts, and after lots of experimentation I like the image much better without the artificial edge enhancement, which also accentuates any noise. I don't think there is any setting for computational precision in FCP. And regardless, FCP was also used for the DV compressed JPEG so the playing field would be level.

Andre, I'm at work so I don't have those images handy to check file size. By all means, don't take my word for any of this though! Run your own tests. This is the chart I used. I'd be interested to see if you got different results using Premiere, Vegas or another NLE.

Oh there is. I am not near my Mac right now, but I am sure you can change the bits per channel somewhere in FCP4. The point is, this might make a difference, quantization noise increases noise, ar at least apparent noise, when you start out with a noisy image. Also I think underexposing in the camera might also have an effect on noise. I would love to do the tests myself, but I don't have a printer, don't have a VX2k, don't have a digicam, except the PDX10 itself... which gives me another idea: you could include in the test the PDX10 in photo mode... and thus see how different the CCD array operates in that mode, or not.

Pondering this some more.... Just looked at my cameras and I think I need to re-shoot those tests one more time (arrgh) because the custom presets were clearly influencing the results. On the last test:

VX-2000
-----------
COLOR LVL = 0
SHARPNESS = -2
WB SHIFT = -4


PDX-10
---------
COLOR LVL = -1
SHARPNESS = -3
WB SHIFT = 0

This explains why the VX-2000 wouldn't white balance (duh!). I'm going to reset the defaults on all these (0). I imagine this will give an image that looks sharper but with stronger white outlines adjacent to the black lines. I don't like the overall feeling of the default sharpness for most things, but maybe it is more appropriate for bringing out detail in wide shots? To be continued...

Please do. We all adore you for the effort you put into it. Don't forget the AE point though. I just can't believe all that noise, and hope that underexposure is part of the reason :-(

Just finished shooting all my tests but I won't get a chance to edit and post until tomorrow, but verrry interesting :

I think these images will look a lot better. I reset all the custom presets to their defaults.

Used the waveform monitor in BTV pro to confirm exposure and focus. What a cool shareware program... I didn't even know it HAD a waveform monitor!

Tom is absolutely right in his frequent comparisons of low light capabilities of these cameras. I used halogen lights indoors with better control this time. The VX-2000 shot at f6.8 0dB while the PDX-10 was f2.8 0dB (according to its data code). That's 2.5 f-stops. Actually, based on the waveforms it's probably more like 2.25 f-stops or something, but the iris doesn't move in clicks that small.

Now this is really interesting: The PDX-10 showed my 16"x12" chart almost exactly full frame. But the VX-2000 showed a white border left and right when the chart filled the frame vertically! That must means the CCD proportions are not correct... when you film with the VX-2000 everything will look a little too tall and skinny! When I get the images into photoshop I can see how many pixels are involved, but it seems significant and is very noticeable. I have never seen any mention of this before. I double and triple checked this to confirm, and used the same tripod without moving it on both cameras. I lined it up vertically and horizontally on the chart as best I could.

I also shot a still photo with the PDX-10. Wow, 4:3 video really discards a huge portion of the CCD. That 690,000 pixels mentioned elsewhere sounds about right.

I re-shot my year-and-a-half old 16:9 tests also since I feel a lot better about exposure and focus on these new tests; I will be sure to export them as high quality JPEG's this time.

I also shot a still JPEG of the chart with my Nikon 5700 and will size that down to 720x480 for comparison. I think it might be better to use this in the tests instead of the raw chart JPEG since the contrast scale on the hardcopy looks different, but we'll see.

Give me a little time and I'll put a whole lot of stuff online, including all the raw images...

It would probably make sense to start separate threads in the appropriate camera forums to discuss these specifics, and I'll followup here with results that relate to the original "wide shot resolution" issue.

Way to go, Boyd! We will be eagerly waiting and Thank You for the work.

After recently shooting extensively with my first DV camera (the DVX100,) I can definitely see how the camera does a better job with close up shots then it does with wide shots. On wide shots everything seems….well…busy…almost as if the focus is soft, but it’s not really the focus. I simply think it’s the limited resolution of SD that is to blame. If I have to use the DVX100 on my feature then I will stick with medium and close up shots as much as possible.

Jim: Yeah, with the small resolution of SD, those wide landscape shots don't look that crisp and detailed. However, there are some ways around it. I have done a test before with a scanned 35mm picture from our Yellowstone vacation. The video (and at that time I was using a 1-chipper Sony TRV-140 Digital8.) didn't look that great on the far off shots, so I took the 35mm pic of the area, scanned it, added a bit of noise/grain to the footage, and put it in w/ the video footage. It looked acceptable to me. Although it was downsampled to 720x480, it looked way better than the video shots of the same area.

You can really go far and do some careful compositing with foreground atmospheric elements w/ motion layered over it, swaying trees, etc, etc... Also, environmental projection mapping would work if you want some camera motion in the shot.

Yeah, that is a lot of work but it may be worth it if you want a high quality shot.

,Frank

The results are in
 

Here's my first attempt at organizing all the test shots I did the other day: http://www.greenmist.com/dv/res

I've tried to include all the tech notes on the webpage itself. The basic format shows 300% enlargements of a small section of the chart for quick comparison. Click on any of these to see the full frame as captured. These frames have not been altered in any way, other than exporting as high quality JPEG's from Final Cut Pro.

What I learned:

(1) The original JPEG file is still by far the cleanest image. The dv-compressed version does show some mosquito noise now however, not quite sure why this is more evident than it was in my earlier test.

(2) When I photographed the chart with my Nikon 5700 and down sampled to 720x480 it also yielded a very clean image with just a slight amount of noise. DV-compressing this image starts getting a bit more ragged, but still far cleaner than either the PDX-10 or VX-2000. Notice how the color and contrast shift due to the limited DV colorspace.

(3) The PDX-10 still is interesting to me. It shows better resolution than the camera in DV mode, but is still pretty noisy (sorry Ignacio ;-) What surpised me was that when I DV-compressed this image it changed very little.

(4) For 4:3 video the VX-2000 clearly beats the PDX-10, but not by a lot. However look at the improved vertical and horizontal resolution and somewhat lower noise level.

(5) Good example of the "warm image" that the Sony cameras are known for. In photo mode the PDX-10 image looks pretty neutral, but switching to video mode produces a pinkish cast. Both examples were manually white balanced. The video image also appears lighter as has been observed elsewhere.

(6) When it comes to 16:9, all those extra pixels on the PDX-10 make a huge difference. No contest here. The DV compression seems especially unkind to the VX-2000 in this example which looks pretty much like a big blob with less than 200 lines vertical resolution.

(7) I repeated the test from http://www.techshop.net/PDX-10/ to divine just how the PDX-10 samples video in the different modes and arrived at similar results, give or take a few pixels.

(8) Probably the most useful thing I learned in this excercise was that you can significantly improve the quality of your DV images if you look at a waveform monitor and adjust exposure and focus accordingly. These new tests are much cleaner than my first attempt.

Conclusions? Well the original question was "why do DV wide shots look so bad?" I only tested two cameras, but from what I saw it seems the answer lies in the cameras themselves. For whatever reason, they don't seem optimized to handle fine image detail. The DV codec itself introduces some artifacts, but nothing nearly as severe as the noisy grain I saw when enlarging framegrabs.

Thanks Boyd for the very interesting pictures, together with the complete picture descriptions. I believe they cover the matter very well. I posted some time ago on the VX2000 uprez issue: a better solution is to shoot in 4:3 and let yr display scaler do the uprez...a significant difference! Not only the difficulties with uprezzing interlaced pics are then gone ( because they first deinterlace before uprezzing) but also because they can uprez to a higher vertical pixelcount than the 525 NTSC lines.

Great work Boyd. Thank You for clearing up the noise issue, so to speak :-)

Boyd, in looking at your test pics of the 4:3 from the PDX10, was there any kind of edge enhancement on ?