Decision Time: Sony SR12 or Canon HF10?

[Ken Ross]

Tom, I don't believe that's how the Sony's 'bias' control works (just a simple raising or lowering of brightness). I believe it is a combination of aperture, shutter speed and gain done in a dynamic way. I know where I've used it I've certainly seen it change shutter speed. Once you get in to a lower ambient lighting conditon and the available shutter speed is 'used up', I'm sure it would revert to aperture and then gain.

But the major difference is that a simple 'exposure' control sets in 'stone' the exposure at that time. It will not vary. Thus, if you zoom in to something after you've set the exposure zoomed out, the picture will darken since your original setting was based on lighting in a 'pre-zoom' condition.

Now if we take that same scenario and apply it to a 'bias' control, the setting would be say 1 stop less (assuming that's how you set it) in your wide angle setting. However the big difference now is that as you zoom, the camera is now adjusting to the reduced lighting as the result of zoom. It will still end up being 1 stop less than what it would have been had you not set the bias control, but it will be correct since the camera is allowed to alter its settings.

[Tom Roper]

Okay, I see what you're saying. The Canon exposure button locks it into manual when used, plus or minus the stops.

When using the bias on the Sony, it would remain in automatic plus or minus the stops.

Does that sound about right?

[Ken Ross]

Originally Posted by Tom Roper

Okay, I see what you're saying. The Canon exposure button locks it into manual when used, plus or minus the stops.

When using the bias on the Sony, it would remain in automatic plus or minus the stops.

Does that sound about right?

Bingo, you got it! So the Sony has the same control as the Canon plus the bias control just as you described it.

[Dave Rosky]

Originally Posted by Tom Roper

Okay, I see what you're saying. The Canon exposure button locks it into manual when used, plus or minus the stops.

When using the bias on the Sony, it would remain in automatic plus or minus the stops.

Does that sound about right?

As Ken has said, this is correct. Sorry for the confusion. I come from primarily a still photography background, and on still cameras (digital or film) the term exposure compensation always means the plus-or-minus bias; locking the current setting is called "AE lock"; and other modes are just called "manual exposure". This is also the case on my older Panasonic AG-50E DV video camera, so I thought the terminology was universal.

I didn't realize that Canon had re-defined the term "exposure compensation" to mean "AE lock plus manual exposure" for their video cameras! No wonder Sony had to call it something else, like AE shift. Too bad Canon did that, resulting in a plethora of different names for the same thing. Consumers aren't already confused enough, I guess ;-)

[Dave Blackhurst]

Got an SR11 in hand, still got lots of testing to do on it. Initial impressions are it's got good ergonomics, and very good build quality. Very nice overall design.

So far I really like the big LCD, interestingly when I switched on guideframe I got little corner brackets that I presume are "safe area" markers - they correspond to the area on my 2.7" screen almost exactly! That said, the recorded video is definitely "wider", showing a wider field of view, but height sure looks about the same amount larger too... so this camera takes in a bit more, that could be nice. Given the wider field of view, the VF is seeming very tiny and inadequate by comparison - good for those times when the LCD wouldn't work, but far preferred the pull out one (HC7/9) to the tilt on this model.

The front control button/knob is also a huge improvement - having seen one on a UX7, was glad to see this - it's not a focus ring, but is better that the wheelie thingy. That you can press and hold and switch functions as needed is pretty handy.

No histogram, but zebras are there, good. Facial recognition seems like a good thing, and looking forward to fiddling with that further, but it appears to be disabled if you are in digital zoom or if exposure is enabled (the latter makes sense given the the FR feature is supposed to auto adjust to faces...)

Now for the initial impressions - I took one still, and hope it's not indicative of the quality - looked like an oil painting, hoping that is an aberration. Lots of patterned artifacts - maybe it was the scene or the flash, but saw them everywhere...

Video looks pretty good, but the CX7 still beats it significantly in low light, with better color and detail - sure wish I knew why the CX7 seems to beat everything else I've tested against in low light, including it's direct siblings - got me puzzled. Comes darn close to being able to focus on a black cat in a dark room. The SR11 was hunting under conditions where the CX could still lock farily well.

SR11 video looks pretty good, but I am seeing some strange color shifts in the blue/green area that I haven't seen in the 7 series, more tests to do no doubt...

[Ken Ross]

Dave, your older model had a larger sensor which probably explained the better low light. But I'm sure you'll see in better light it will look a lot nicer than the CX7.

By the way, I'm sure you've seen me say a number of times that the still quality is in no way indicative of the video quality. But with that said, using the "photo" button, I've gotten some very impressive, detailed shots in 1920X1080 that look great on my 60" Pioneer plasma.

[Jason Ivins]

I've been very impressed with still pictures taken in good light. Good enough to leave my point and shoot at home.

In poor light or when the flash is needed, the still pictures leave a lot to be desired. Red eye is a huge problem given the location of the flash. I'm sure an add-on flash would improve this.

[Dave Blackhurst]

There's not a whole lot of difference between the CX7/HC7 sensor and the SR11 sensor in size, but the CX7 even beats the HC7/9 (same sensor) by a small margin - it's cleaner and better colors in low light conditions. Got me puzzled, but it's there...

I bought the CX7 for that reason, as the specs stated it was better, and it's proven to be the least noise and best low light performance so far - we're talking split hairs, but something is going on there.

I may post two short clips of a low light indoor test I ran... not perfectly side by side, so I should run it again "scientifically", but the CX is slightly cleaner - I also discovered that the CX7 LCD "underscans" (overscans?!? I think that's the right one, I get it mixed up). I set the cams up so that I was framing based upon the SR11 brackets and the edge of the CX7 screen - surprisingly when dropped into Vegas, the height and width actually showing were identical - the CX7 LCD "crops" the video on all sides so you actually get more than you see on the LCD, where the SR11 shows more accurately what you will get...

I shot some outdoor stuff, but when I dragged all but the first clip into vegas it locked up - looks like the latest PMB software is needed to view ANY clips with this model, though I need more tests... I'm not sure that having to have a software interface to use a camera is a step ahead - firewire/miniDV were more universal, but these new cams mean don't lose the disk...or else...


As for the "still" function, I'm actually testing the "still" function of the camera, which as advertised is significantly higher res than a 1920x1080 screen grab... and so far, I'm seeing some pretty ugly artifacting in the stills from this cam, both at the "10.2M" and the "7.2M" it shoots while video is running (it does however shoot stills fairly continuously, and I will be testing with a MS Duo later - with video running it sometimes has to wait for the still to become available again). I'm sure the stills would look OK in 4x6 or 5x7, but when you zoom in on them (or if you blew them up larger), I'm not sure how the results would turn out - I've had a fair amount of experience cleaning up stills, and these have some things going on that don't filter out... the CX7 stills (6.1 & 4.6M) show no such noise or artifacting - in fact they are very clean, as are stills from the HC7. I'll take some side by side and post, but there is a distinct artifact pattern that is NOT something that I expect - it was actually a big surprise to see what I can only describe as an "oil painting" type of look to the stills - it's noise, not detail.

Not to worry, I'm not going to compare apples and oranges - it will ultimately be the still function of one cam vs. the still function of the other, same for video...

I happen to find the dual use capability handy, so the quality of that function actually has some value for me. Vegas pulls great stills off the timeline, but it would be nice to have a higher res file to work with.

I'll run more tests, but so far I'm not seeing any big jump in quality over the "older" cameras, which is a bit of a surprise. Motion is one thing I'll be testing for if the wind keeps up tomorrow, though I've had no issues with the CX7 in that department either...

I'm not going to knock the SR11, in fact it's got a LOT to like about it, I'm just seeing it as an incremental evolution of some already excellent cameras, and I'm nitpicking where most anyone would probably never notice. And I'm going to qualify these observations as "early impressions".

[Dave Blackhurst]

Originally Posted by Jason Ivins

I've been very impressed with still pictures taken in good light. Good enough to leave my point and shoot at home.

In poor light or when the flash is needed, the still pictures leave a lot to be desired. Red eye is a huge problem given the location of the flash. I'm sure an add-on flash would improve this.

As long as the pictures are clean, redeye removal in post is no big deal. So far, I'm happier with the stills from my older HC7 and CX7, both of which have been excellent for stills, but more tests are needed... got to shoot side by side and see what happens - apples to apples.

SOME of the stills seem pretty good, but they all defnitely have this "oil painting" look when you zoom in - I've never seen anything quite like it in ANY camera I've owned, and the photo retouch programs I've used to retouch sucessfully don't know what to do with it... and this includes a few shots in excellent lighting conditions.

I've just pulled up a bunch of stills from a variety of cameras both still and video/dual mode, and the stills from the SR11 are the ONLY ones that exhibit this noise/artifact pattern. I'm taking a sampling of nearly a dozen cameras - everything from VGA res stills up to 10.1M shots from a Sony R1...

[Florin Andrei]

Originally Posted by Justin Halliday

Sony SR12
Pros:
120GB HDD

HDD? I thought that's in the Cons list. :-)

[Ken Ross]

Originally Posted by Jason Ivins

I've been very impressed with still pictures taken in good light. Good enough to leave my point and shoot at home.

I was actually surprised how good still outdoor shots look on a 60" plasma. In fact, it does look like a high quality freeze frame from full HD video. Pretty impressive actually. Unlike Dave, I'm seeing zero noise, I mean no noise at all when viewed on my 60" 1080p plasma. But these were all outdoor shots.

I don't ever see using a video camera to replace a dedicated still camera, that approach is always going to be a compromise. It's really no different than taking your digital still camera with 'video function' and expecting to get great results taking video...you won't. One camcorder may be somewhat better or worse than another camcorder for a picture taking function, but all will pale by comparison to a good digital still cam.

[Ken Ross]

Originally Posted by Dave Blackhurst

I shot some outdoor stuff, but when I dragged all but the first clip into vegas it locked up - looks like the latest PMB software is needed to view ANY clips with this model, though I need more tests... I'm not sure that having to have a software interface to use a camera is a step ahead - firewire/miniDV were more universal, but these new cams mean don't lose the disk...or else...

Dave, sounds like you're doing something wrong or are perhaps missing a codec. I can play the SR12 clips in virtually any media player or editing program. Programs I've tried and that have worked include: Windows Media Player, Windows Media Classic, Vegas, Canopus Edius and ULead Studio 11+. I have the Core AVC Codec which I highly recommend if you're going to get serious with AVCHD. You can download it for around $15.

Originally Posted by Dave Blackhurst

I'll run more tests, but so far I'm not seeing any big jump in quality over the "older" cameras, which is a bit of a surprise. Motion is one thing I'll be testing for if the wind keeps up tomorrow, though I've had no issues with the CX7 in that department either...

My only testing with the CX7 was in a couple of Best Buys with memory stick in hand. In my testing the SR12 was significantly better under those conditions. I found the CX7 to be somewhat soft under those conditions...certainly relative to the SR12. But I never got to test a CX7 outdoors.

By the way Dave, even though the CX7 does not have a significantly larger chip than the SR11/12, it's probably enough of a difference to show a bit better quality in low light. The chip size numbers may not seem that different, but keep in mind these sensors start out really small to begin with, so sometimes a small difference may be larger than you think. But with that said, the SR12 holds up very nicely in low light to my HV20...better at times in fact and the HV20 has a pretty good low light reputation (at least for HD cams which are actually all pretty poor). :)

[Ken Ross]

Originally Posted by Florin Andrei

HDD? I thought that's in the Cons list. :-)

If you're looking for huge internal storage for almost non-stop shooting, it's a 'plus'. If you're looking for maximum shock resistance, it's a 'con'. ;)

[Bruno Donnet]

Originally Posted by Ken Ross

I agree Tony, it would have been nice if they gave you that adjustment. Even though I've always kept the adjustments at default in my HV20, it's nice to know they're there. On the other hand the Sony gives you the ability to adjust 'bias' from what the cam wants to do. So it will constantly adjust exposure up or down a pre-set amount from what it would have done alone. This is in addition to being able to adjust exposure for just one shot like the Canon. The Sony allows the same thing with white balance, continuously adjusting warmer or cooler from what it would have done alone. These are features I'd like to see the Canon have.

I'm not fully agree with you, Ken, the Canon HV20/HV30 (and I presume the HG10, HF10/HF100) have some 'bias' adjustments for the exposure, color, contrast and sharpness.

Only the latitude of each set is more limited: instead of the possibility to adjust from -4 to +4 on the Sony consumer camcorders, the Canon cams offer only -1 to +1 (so only 3 choices: less, default, more!).

With my past Sony HC1, I've in fact never used all the latitude proposed by Sony because when you select more than 2 ticks in - or in +, the result is totally unusefull. So the only -1 or +1 on the Canon seem a little frustating, but, in fact... are enough in 90% of the cases.

I don't remember the exact setting of all these adjustments on my own HV20, but I thing that at minimum the exposure (named 'Brightness' in the menu) is at -1 to restrict as much as possible the oversatured whites, and, if I remember well my test on the colors, the color setting is at -1 too (giving a little cooler image that matches better my HC1 videos that have been shot at 'Color shift' +1 on the Sony side).

For sure, as I said already, the possibilities are more limited, but, you cannot say that the Canon cams have no 'bias' adjustments at all.

The other difficulty is to found the right entry point into the Canon menu (espacially when you use, or have used, a Sony cam): it's in mode 'P', under 'Image effect' --that's confusing with what is 'Image effect' on a Sony cam...--, and under 'Custom'.
The mode 'P' ('programmed' mode) doesn't mean that you loose the automatisms: all the other settings can stay on their Default value or in Auto mode: theses 'Custom' adjusments work really as 'bias'.

[Ken Ross]

Bruno, the limited adjustment of the Canons (brightness, contrast etc.) are not the same as an exposure adjustment. Exposure controls aperture, shutter speed and gain. I don't beleive the Canon works that way by controlling brightness and contrast.