DSR-450WS vs. SDX 900 and DVX 100A

[Jeff Regan]

I finally got the chance to look at my Sony DSR-450WS next to a Panasonic SDX 900 and DVX 100A. I wanted to make scene files that emulated the colorimetry of the two Panny cameras. I also got a chance to test the P+S Technik Pro 35 lens adaptor and a Zeiss Super Speed prime.

 

I used matching Fujinon 15X8 lenses on the 450 and 900, a MacBeth color chart, plus chips, resolution and

a fleshtone chart. I put an overlay on a vectorscope to plot the "dots" in order to help match with the matrix

on the 450. This was not a full-on camera comparison, I didn't rate the cameras sensitivity wise or do a

real latitude comparison, as my priority was coming up with some camera emulation scene files.

 

Here are some general observations:

 

The 900 and 450 were very clean and quiet SN wise, the 100A was noisy.

 

The base colorimetry with matrix on of the 450 and 900 are pretty similar, as both cameras are designed to

ITU-709 worldwide colorimetry specs. There is an overall difference between the Sony and the two Panny's,

the Sony being more to the warm/magenta vs. the Panny's with a more blue/green bias.

 

I put the 900 into Scene 4 Filmlike mode and matched my 450 using the user matrix. I got pretty close,

the 900 has good red saturation and very strong in the cyan spectrum.

 

The DVX 100A was used in Scene File 5 mode, which is a 24P film like mode. It also had strong red saturation and VERY strong green saturation, more than I could add via the 450 matrix range. It was not as plus cyan as the 900 in Scene 4.

 

Comparing exterior images between the three cameras of wide shots of homes/cars/foliage, the 450 looked

the sharpest with detail off or on with all cameras having reduced detail for their film modes. This was confirmed with a resolution chart. The 450 had more shadow detail than the 100A with film gamma 1 using black gamma on in mid-high mode, even though the 100A gamma crossover was at 60IRE vs. 55IRE for the 450.

 

On foliage, the 900 made all the leaves on a hedge one shade of green, whereas the 450 showed tonal

differences, with some leaves being more yellow than others. The 100A, even with its strong green push,

showed more shades of green on the same bush than the 900.

 

Because the 450 has twice the pixel count of the Panny CCD's, it actually produced more artifacts in 24P mode on fine detail, such as window shutters, car grills, shingles. There was line twitter and moire that

didn't show on the Panny's. The 450 resolves more detail than the NTSC format can deal with.

 

On a fleshtone chart, the 100A made a pleasing, saturated rendition of the little girl's face, but the 450

showed more natural detail, such as freckles and tonality. After matching the 450 to the 100A using a

MacBeth chart, then shooting the flesh chart, the 100A still looked more saturated, but with less detail

and tonality. I had to boost R-G and B-G on the 450 matrix to get a similar fleshtone saturation, but

that skewed the MacBeth color chart match.

 

I found latitude in the highlights to be very similar for the 450 and 900, and the 100A was better than I

expected, although it didn't have the shadow detail--it looked more contrasty and saturated, less subtle

in gray scale gradiation.

 

The 900 has the 4:2:2 color space advantage going for it over the other two cameras, the 100A is an amazing camera for the money, just too noisy compared to the super clean 2/3" cameras, the 450 has more natural resolution than any SD camera I've seen, but this causes artifacts on very fine detail--particularly when monitoring in composite. I didn't have enough time to really get to know the Panasonic cameras in enough different conditions, but I at least got an idea of what they are capable of.

 

Finally, with my testing of the Pro 35 adaptor and 65mm Zeiss Super Speed, I found the system to be a good match for the DSR-450WS, as there is some slight softening of the image. The difference in depth

of field is startling vs. a normal 2/3" video lens.

 

Jeff Regan

Shooting Star Video

www.ssv.com

[Peter J DeCrescenzo]

... Comparing exterior images between the three cameras of wide shots of homes/cars/foliage, the 450 looked the sharpest with detail off or on with all cameras having reduced detail for their film modes. This was confirmed with a resolution chart. ... Because the 450 has twice the pixel count of the Panny CCD's, it actually produced more artifacts in 24P mode on fine detail, such as window shutters, car grills, shingles. There was line twitter and moire that didn't show on the Panny's. The 450 resolves more detail than the NTSC format can deal with. ...

 

Hi Jeff: Thanks for the interesting report.

 

Depending on the contents of a scene, I've also noticed slight detail-induced line twitter such as you mention, even when the cam's detail is set low or off.

 

I wonder if you or others might recommend a filter which produces a subtle amount of diffusion -- just enough to smooth fine details before they hit the cam's CCDs?

[Michael Nash]

Thanks for the detailed comparisons (no pun intended).

 

When shooting with the SDX and trying to emulate a Sony ENG look, I ended up raising the red gamma to warm it up. Totally by eye and on the fly, based on my experience with Sony cameras. Using the color matrix is of course much more accurate and doesn't give you a hue shift throughout the luminance range.

 

That "twitter" isn't really the result of too much information, it's more like a "mismatch" between the camera's sensor and the system. But either way it's still there, and has to be dealt with. One thing you can do is adjust the detail frequency, which controls the pixel width of the enhancement line. I usually try to go for as high a frequency as possible (thinnest line) in SD so that I can still get some of the benefits of enhancement, without so much of the cartoony outline look.

 

In my experience moire and twitter problems are rarely solved via the detail though. Usually if the problem's there at all it's the result of the pixel mismatch and detail doesn't change it much. For filters you could try low grades of Tiffen SoftFX, Schneider Classic Softs, or Mitchell Diffusion. All of those blur fine detail without changing contrast or halation much. But remember that a camera's detail enhancement is electronically applied to the signal AFTER the filter and CCD, and you can end up putting edge detail back on top of an optically softened image. You have to adjust both the filters and detail in tandem to get the right look.

 

A nice follow up test might be to compare the default and maximum settings of master gamma, knee, and black stretch. See how they look "right out of the box," and then push them to their limits to see what they can do and where they fail (dynamic range, noise, color artifacts, and so on).

[Jeff Regan]

Hi Jeff: Thanks for the interesting report.

 

Depending on the contents of a scene, I've also noticed slight detail-induced line twitter such as you mention, even when the cam's detail is set low or off.

 

I wonder if you or others might recommend a filter which produces a subtle amount of diffusion -- just enough to smooth fine details before they hit the cam's CCDs?

 

Hi Peter,

 

I've used Tiffin Pro-Mist and Black Pro-Mist filters for years, 1/8 to 1/4 would be a subtle amount of diffusion.

I also like Tiffin Soft FX and Diffusion FX filters in the lowest grades. I used to shoot a lot with very shear nets behind the lens, but this is not a subtle effect. I'm not sure if all of the artifacts I saw would be totally tamed by these diffusion filters, but they would help.

 

I've got my filter inventory posted on my site and you're welcome to come by sometime and test filters until

you drop!

 

Jeff Regan

Shooting Star Video

www.ssv.com

[Jeff Regan]

Thanks for the detailed comparisons (no pun intended).

 

When shooting with the SDX and trying to emulate a Sony ENG look, I ended up raising the red gamma to warm it up. Totally by eye and on the fly, based on my experience with Sony cameras. Using the color matrix is of course much more accurate and doesn't give you a hue shift throughout the luminance range.

 

That "twitter" isn't really the result of too much information, it's more like a "mismatch" between the camera's sensor and the system. But either way it's still there, and has to be dealt with. One thing you can do is adjust the detail frequency, which controls the pixel width of the enhancement line. I usually try to go for as high a frequency as possible (thinnest line) in SD so that I can still get some of the benefits of enhancement, without so much of the cartoony outline look.

 

In my experience moire and twitter problems are rarely solved via the detail though. Usually if the problem's there at all it's the result of the pixel mismatch and detail doesn't change it much. For filters you could try low grades of Tiffen SoftFX, Schneider Classic Softs, or Mitchell Diffusion. All of those blur fine detail without changing contrast or halation much. But remember that a camera's detail enhancement is electronically applied to the signal AFTER the filter and CCD, and you can end up putting edge detail back on top of an optically softened image. You have to adjust both the filters and detail in tandem to get the right look.

 

Hi Michael,

 

I agree with your observations. I remember when the Sony BVP-90 first came out in the early '90's.

Clients and DP's used to the look of a BVP-70IS or BVP-7 thought the BVP-90's were soft. That was

because the BVP-90 detail circuit worked at higher frequencies which made the enhancement more

thin and subtle. Sony actually had to come out with a new detail board that worked like the older

enhancement circuits, I remember having both boards for my BVP-90. I hate the look of over

enhanced video cameras, but many equate artificial enhancement with resolution and real detail.

 

On the DSR-450WS, the artifacts that I saw do not go away with detail turned off, just like you said.

They become more subtle in some instances, but they are there. What I like about the 450 is that

it is the first SD camera I have ever owned that still has adequate natural detail and sharpness with detail

off.

 

The majority of the artifacts that I see, and only on certain kinds of scenes, are related to progressive

modes, not interlace, but these progressive artifacts were not present on the Panasonic cameras.

 

Jeff Regan

Shooting Star Video

www.ssv.com

[Peter J DeCrescenzo]

Hi Peter, I've used Tiffin Pro-Mist and Black Pro-Mist filters for years, 1/8 to 1/4 would be a subtle amount of diffusion. I also like Tiffin Soft FX and Diffusion FX filters in the lowest grades. ... I've got my filter inventory posted on my site and you're welcome to come by sometime and test filters until

you drop!

 

Thanks for the info, Jeff & Michael. And Jeff, I may be in touch with you sometime soon concerning your very generous offer. :-)

[Jac Chesson]

Jeff,

 

Thanks for going to all that trouble and posting your findings. It was very helpful and educational.

 

I personally love my 450, but have noticed the twitter problem and would love an in-camera solution to it. I wonder if the folks at Sony could come up with something.

 

I need to send my 450 in to Teaneck, NJ (to have them fix my timecode preset feature!) and maybe I'll bring up the twitter issue then. They seem to be very helpful and responsive.

 

Thanks, everyone! This is the only place I've found that actually talks about using our camera. It's great!

Still would love to see an exclusive Sony 450 forum, but I don't think the numbers are in my favor!

 

Jac

[Peter J DeCrescenzo]

... This is the only place I've found that actually talks about using our camera. It's great!

Still would love to see an exclusive Sony 450 forum, but I don't think the numbers are in my favor!

Yeah, all those RED camera owners really out-number us! B)

[Ralph Oshiro]

Thanks, everyone! This is the only place I've found that actually talks about using our camera. It's great! Still would love to see an exclusive Sony 450 forum, but I don't think the numbers are in my favor!

As a recent DSR450 owner, I agree! Maybe cinematography.com could find it in its heart to give a home to our poor, orphan camera of choice? As a former/current Panasonic DVX100 owner, I used to frequent DVXUSER.com quite often (and still do), but Jarred has chosen to refocus the site to become more DVX/HVX-centric (as well as he should). But, WE NEED A HOME! Cinematography.com seems like a good fit! Anyone know the moderators here personally?

 

Regards

 

 

Jeff:

 

Great post! It's been so long since I've posted here, that I forgot my old username and had to re-register! Look forward to more of your findings, Jeff!

 

[Rezfactor]

[DMW]

I finally got the chance to look at my Sony DSR-450WS next to a Panasonic SDX 900 and DVX 100A. I wanted to make scene files that emulated the colorimetry of the two Panny cameras. I also got a chance to test the P+S Technik Pro 35 lens adaptor and a Zeiss Super Speed prime.

 

I used matching Fujinon 15X8 lenses on the 450 and 900, a MacBeth color chart, plus chips, resolution and

a fleshtone chart. I put an overlay on a vectorscope to plot the "dots" in order to help match with the matrix

on the 450. This was not a full-on camera comparison, I didn't rate the cameras sensitivity wise or do a

real latitude comparison, as my priority was coming up with some camera emulation scene files.

 

Here are some general observations:

 

The 900 and 450 were very clean and quiet SN wise, the 100A was noisy.

 

The base colorimetry with matrix on of the 450 and 900 are pretty similar, as both cameras are designed to

ITU-709 worldwide colorimetry specs. There is an overall difference between the Sony and the two Panny's,

the Sony being more to the warm/magenta vs. the Panny's with a more blue/green bias.

 

I put the 900 into Scene 4 Filmlike mode and matched my 450 using the user matrix. I got pretty close,

the 900 has good red saturation and very strong in the cyan spectrum.

 

The DVX 100A was used in Scene File 5 mode, which is a 24P film like mode. It also had strong red saturation and VERY strong green saturation, more than I could add via the 450 matrix range. It was not as plus cyan as the 900 in Scene 4.

 

Comparing exterior images between the three cameras of wide shots of homes/cars/foliage, the 450 looked

the sharpest with detail off or on with all cameras having reduced detail for their film modes. This was confirmed with a resolution chart. The 450 had more shadow detail than the 100A with film gamma 1 using black gamma on in mid-high mode, even though the 100A gamma crossover was at 60IRE vs. 55IRE for the 450.

 

On foliage, the 900 made all the leaves on a hedge one shade of green, whereas the 450 showed tonal

differences, with some leaves being more yellow than others. The 100A, even with its strong green push,

showed more shades of green on the same bush than the 900.

 

Because the 450 has twice the pixel count of the Panny CCD's, it actually produced more artifacts in 24P mode on fine detail, such as window shutters, car grills, shingles. There was line twitter and moire that

didn't show on the Panny's. The 450 resolves more detail than the NTSC format can deal with.

 

On a fleshtone chart, the 100A made a pleasing, saturated rendition of the little girl's face, but the 450

showed more natural detail, such as freckles and tonality. After matching the 450 to the 100A using a

MacBeth chart, then shooting the flesh chart, the 100A still looked more saturated, but with less detail

and tonality. I had to boost R-G and B-G on the 450 matrix to get a similar fleshtone saturation, but

that skewed the MacBeth color chart match.

 

I found latitude in the highlights to be very similar for the 450 and 900, and the 100A was better than I

expected, although it didn't have the shadow detail--it looked more contrasty and saturated, less subtle

in gray scale gradiation.

 

The 900 has the 4:2:2 color space advantage going for it over the other two cameras, the 100A is an amazing camera for the money, just too noisy compared to the super clean 2/3" cameras, the 450 has more natural resolution than any SD camera I've seen, but this causes artifacts on very fine detail--particularly when monitoring in composite. I didn't have enough time to really get to know the Panasonic cameras in enough different conditions, but I at least got an idea of what they are capable of.

 

Finally, with my testing of the Pro 35 adaptor and 65mm Zeiss Super Speed, I found the system to be a good match for the DSR-450WS, as there is some slight softening of the image. The difference in depth

of field is startling vs. a normal 2/3" video lens.

 

Jeff Regan

Shooting Star Video

www.ssv.com

 

 

 

Ohh! What a report! I am amused!!! Thanks Jef to share the report with us. It helps me a lot for future work. Does anybody have similar report or findings with Sony DSR 400 and JVC GY HD 100U? Please share in this thread or kindly open a new thread.

 

Thanks once again.

 

www.adexpressions.com

[Ralph Oshiro]

SONY DSR450 FORUM PLEA:

 

Well, guys . . . I just sent a lengthy plea to the forum administrator here, Tim Tyler, BEGGING, for a Sony DSR450 forum here at cinematography.com. Perhaps a few more PMs (and maybe even promises of future multiple paid subscriptions!) will entice Tim to see things our way. There are some awfully knowledgable people here at cinematography.com, and I think that this site is a PERFECT fit for Sony's version of the much-heralded SDX900, Sony's DSR450WS!

 

Regards,

 

Ralph Oshiro

 

Proud new owner of the lowest-cost, highest-quality, 24P SD indie-cinema camera, the Sony DSR450, the camera without an online discussion forum to call its own . . .

Edited June 15, 2006 by Ralph Oshiro

[Simon Wyndham]

Good to hear you're liking your new camera Ralph! See what I mean about it producing a detailed picture with the detail off? :)

[Guest TVTim]

Hi Guys

 

I too would like to see a DSR450 section here. It is so hard trying to find 450 info on the web. It is mostly saturated with best deals and shop advertisements for DSR450's.

 

Tim.

[Stephen Edwards]

Hello Jeff,

Thanks for the in depth report on the DSR-450! I am learning so much about my advanced camera set up through this forum. Pages of notes have been taken down comparing everones very kind expert tech spec advice and am now ready to start "setting up" the camera to the way I shoot and to my own particular tastes.

I also got a set of 82mm tiffen's; warm uv and polarizer to play with as I dial my camera in. Had to reduce the auto iris 1/2 stop and will set my white clip to 100 among many other adjustments since I only do TV Video work.

Shot the sunset the other night on -3 and boy did it look clean all the way til dark!

Feeling much better about my purchase every day thanks to all you guys...