GH1, meet the DSC MegaTrumpets res chart

Thanks for these tests Barry. To me the Gh-1 seems to be the better choice of it & the 7D , because you get similair performance and a decent lens, while the Canon gives you a nice DSLR body. These companies are getting close to the answer but No Cigar yet.

Unfortunatley I'm still where I was 6 years ago,
Can someone please make a Video camera with the proper size sensor, detachable lens, affordable price, good mics pres, limiters, good codec, etc... we laid out the criteria years ago... c'mon people is it really that tough? Will it take special brain surgery on the deciding execs for these companies to get it right?

I will start a donation fund for their labotomies if they promise to deliver what we stated afew years back.
​
RED was going to be the answer and then they went off into outerspace with all this high performance dragcar racing stuff (really good but it ups the cost and delays delivery) so it doesn't appear to be the long sought after answer we have all been looking for.
 
Thanks Barry. Pretty much confirmed that neither is exactly where I want to be, but for the form factor, the lenses, the dof, and the learning curve I want a DSLR and the codec alone plus the possible downsides are not going to justify scrapping my GH1 and making an effort to get a 7D.

Just a question about true hi-def, etc. Would a truly HD camera resolve around 900 lines in theory? And what do mini-dv cameras like the DVX come in around, ballpark?

Thank you again!
 
Ben_B said:
By sharper I didn't mean to say anything about in camera sharpening, I was referring to the way in which they seemed to be resolving the edges of the patterns...the 7D seemed a little fuzzier, the Gh1 seemed sharper/harsher.
Click to expand...
But what I meant was, those factors weren't specifically equalized or attempted to be equalized, so ... any judgements wouldn't necessarily be valid. That's all.

What lens was this done with with each camera?
Click to expand...
Zeiss ZF 50mm f/1.8, stopped down to 5.6.

That Zeiss is nasty wide-open... :zombie_smiley:
 
10s said:
RED was going to be the answer and then they went off into outerspace with all this high performance dragcar racing stuff (really good but it ups the cost and delays delivery) so it doesn't appear to be the long sought after answer we have all been looking for.
Click to expand...
The Scarlet may indeed be what we're looking for; it kind of depends on the price tag. If it's $7500 for a brain only, that will rule it out for many folks. I doubt you're going to get much better than what we've got, for $1500-$1700! I mean, consider that the GH1, if sold as body-only, would be $800. That's 1/10 the price of the Scarlet with a comparable sensor size. Will the Scarlet be worth 10x more? If it performs like I'm now thinking it will, my answer will be "yes", but for a lot of people it will be too much of a stretch for their wallets.

Now, let's put it in perspective -- yes, I'm convinced that these cameras are "faking it" in terms of the detail they're presenting. But, the images look fan-friggin'-tastic. Yes you can get more legit imagery, sharper true images, real detail, and no aliasing -- but at what price?

Know what the limitations are, work within them, and these little DSLRs can do amazing work.
 
Kellar42 said:
are not going to justify scrapping my GH1 and making an effort to get a 7D.
Click to expand...
I gotta tell you, the 7D feels like a primitive hunk compared to the sleek little GH1. I forgot just how tiny and slick that thing is. I like them both, but... the GH1 just feels more like a sports car, the 7D feels like a pickup truck.

If it wasn't for the codec puking, the GH1 would be head and shoulders the preferred camera for most DVXUsers, in my opinion. It's soooo close, but the codec and the stupid pulldown introduce issues that the 7D solves. The 7D has issues the GH1 solves.

Again, all these companies are comparably competent, and within a given budget each is going to produce a competitive product. You just really have to decide which set of compromises suits you best, because at $1700, there's gonna be compromises.

Just a question about true hi-def, etc. Would a truly HD camera resolve around 900 lines in theory?
Click to expand...
The HPX300 and EX1 both resolve 800 lines cleanly, and that's the maximum the charts I used with them went up to. They were resolving more than the charts could show.

And what do mini-dv cameras like the DVX come in around, ballpark? Thank you again!
Click to expand...
A standard-def DVX can pull about 540 lines in 4:3 mode, and about 360 lines in Squeeze mode.
 
IMO they are both deliberately crippled. Pretty much what we were expecting all along but its
great to have confirmation. Sorry if its sounds very neg but its true. This is the first round of 24p cmos after all.

BTW - Great analogy with pickup/sports car and a big thank you Barry for doing all of these tests.
 
Last edited:
I don't think they're "deliberately crippled", I just think they're $1500-$1700. One tenth the price of a Red. One 4th the price of an EX1. You can't get something for nothing. They do what they can within the budget they've got.

I've seen a chart from the Pentax K-7, and it looks pretty much the same as these. It's what you can get from the technology available at this incredibly cheap bargain-basement price bracket.
 
To me AGC is a deliberate cripple. Just think, six years ago I bought a little $300 palm sized canon camcorder (ZR65 I believe) that was standard def mini DV and it had a switchable mic/line in with no agc. Sorry to disagree with you Barry but IMO a deliberate cripple is the only way to see it.

Now I know I illustrated this example with audio but it is the biggest red flag in the cripple department so I chose that. My point is that once you recognize that a company has ethics that deliberately cripple a cams audio system you should therefore logically assume they could and would cripple other systems as well - including image quality. 7D & GH1 ... I own these cams, love these cams, use these cams but I won't defend them.

Both pana and canon could have put out better cams but they make more money when they pull their punches. But thats old ground. I'll shut up now.
In a grumpy mood, must be the ad agencies I work for.

EDIT:
apologies once again for my mood swings, Barry I have the utmost respect for you and even greater respect for your
constant even keel positive attitude. Plesae don't mind me.
 
Last edited:
Barry_Green said:
I'm convinced that these cameras are "faking it" in terms of the detail they're presenting. But, the images look fan-friggin'-tastic.
Click to expand...

But what effect does turning sharpening off with both cams? That might reveal when shooting this chart how much of the fake is binning and how much in camera processing.
 
J Davis said:
Now I know I illustrated this example with audio but it is the biggest red flag in the cripple department so I chose that. My point is that once you recognize that a company has ethics that deliberately cripple a cams audio system you should therefore logically assume they could and would cripple other systems as well - including image quality.
Click to expand...
What Barry was saying was that at $1700 and below, you can't expect these cameras to be properly designed for serious HD video. The video feature is a hack on these still cameras. The image quality issues are a consequence of that fact. So it's not like these companies decided, "Oh, let's protect our camcorders by worsening the video image quality on the DSLRs."

Nikon has no camcorders to protect, and its video DSLRs have even worse issues. And those issues come from the same reason above: Video is a hack.
 
But what effect does turning sharpening off with both cams?
Click to expand...
You can't turn off sharpening with the GH1 - at -2, it's still sharpening, but less than at -1.

I can't help wondering what result would be obtained with the GH1 stock lens. On the face of it, the test could simply be revealing the limitations of the lens used (I realise the lens is actually probably not the limiting factor, but...) especially as it's being compared in the commentary with other devices using their stock lenses, not the lens under test.

I don't think I added my voice to the "thanks for troubling to do these tests" chorus before. I'm singing with the choir as loudly as anyone else, honest.
 
stephenvv said:
But what effect does turning sharpening off with both cams? That might reveal when shooting this chart how much of the fake is binning and how much in camera processing.
Click to expand...
No difference. The lines get grayer when sharpening is turned off, and more distinctly black and white when it's turned all the way up. On the 7D thread I posted a pic of what it looks like with sharpening all the way off. It doesn't affect the aliasing one bit.
 
Ozpeter said:
I can't help wondering what result would be obtained with the GH1 stock lens. On the face of it, the test could simply be revealing the limitations of the lens used (I realise the lens is actually probably not the limiting factor, but...) especially as it's being compared in the commentary with other devices using their stock lenses, not the lens under test.
Click to expand...
The lens makes no difference on the video downrez version. The still images are a million times sharper and cleaner with no aliasing whatsoever, when shot with the stock lens. I did that test on the 7D, because people complained that I was using the stock lens (sigh). So if I use the stock lens, they complain that I used the stock lens and if I use a better lens, they complain that I didn't use the stock lens... :)

Anyway, the point is, if you take a still shot of the chart, the detail is absolutely screamingly luscious and the chart rendering is nearly perfect. Here's a shot from the stock lens of the 7D, shot at still-camera resolution and downconverted to 1080P in PhotoShop:
MegaTrumpets-Still-Cam.jpg


Compare that to the video chart and you can see that the issues and ugliness introduced in the image are absolutely nothing to do with the lens. The lens is capable of producing the chart beautifully.

I have no doubt the GH1's stock lens would perform comparably, I just didn't happen to take a shot with it at the same time. I have an ISO 12233 home-printed chart that I can show though, it was shot with the stock lens, so I think should suffice:
GH1-ISO12233-extraction-stock-lens.jpg


That's clean and pristine detail down to around 1400-1500 lines. No aliasing, no contamination. So, in summary -- the stock lens is way more than capable of handling the highest demands of 1080p video, and is probably 4x as sharp as HD video would require.
 
Barry_Green said:
No difference. The lines get grayer when sharpening is turned off, and more distinctly black and white when it's turned all the way up. On the 7D thread I posted a pic of what it looks like with sharpening all the way off. It doesn't affect the aliasing one bit.
Click to expand...

I went and found that on the 7D thread - thanks. But although your assumption that the GH1 will be the same may be right, it's also clear from the test it does not use the exact same method as the 7d, so it's possible with sharpening turned off it's results may differ.
 
But although your assumption that the GH1 will be the same may be right, it's also clear from the test it does not use the exact same method as the 7d, so it's possible with sharpening turned off it's results may differ.
Click to expand...
I shot it with sharpening turned off, and it doesn't make a difference. Had it made a difference I would have posted that photo.
 
ydgmdlu said:
The video feature is a hack on these still cameras.
Click to expand...

Barry_Green said:
Anyway, the point is, if you take a still shot of the chart, the detail is absolutely screamingly luscious and the chart rendering is nearly perfect. Here's a shot from the stock lens of the 7D,
Click to expand...

A new day and seeing things clearer now. I still think agc is a deliberate cripple but ydgmdlu is right about image, video is a hack. Barry's still brings this home with finality.
Amazes me that we all flock to a hack.
 
J Davis said:
A new day and seeing things clearer now. I still think agc is a deliberate cripple but ydgmdlu is right about image, video is a hack. Barry's still brings this home with finality.
Amazes me that we all flock to a hack.
Click to expand...
Flock to, and love it. The price tag is the main thing. If this was $30,000 I don't think we'd be nearly so thrilled with it. But at $1500, it's insanely good. Even if all the "fine detail" is actually aliasing artifacts! :thumbsup:
 
Barry_Green said:
I shot it with sharpening turned off, and it doesn't make a difference. Had it made a difference I would have posted that photo.
Click to expand...

Thanks - but seems odd. There should be more aliasing at sharpening max than sharpening min if this is a sensor binning effect as the sharpening would be applied post-binning.

The question I'm dancing around here, in real-world shooting of fine (likely to alias) detail, does shooting with sharpening off give the best (or least bad) result?
 
It doesn't change aliasing at all. Aliasing is happening at the sensor-scanning level. Sharpening is applied afterwards.

If you want accurate images, you should always shoot with sharpening off anyway, even on imagers that aren't subject to aliasing... :thumbsup:
 
You must log in or register to reply here.
Back
Top